Carbohydrate Chemistry

Specialist Periodical Reports
Edited by A Pilar Rauter
Carbohydrate Chemistry
Chemical and Biological Approaches
Volume 38
Chemical and Biological Approaches
Volume 38
A Specialist Periodical Report
Chemical and Biological
Approaches
Volume 38
A Review of the Literature Published between
January 2011 and February 2012
Editors
Amelia Pilar Rauter, Universidade de Lisboa, Portugal
Thisbe K. Lindhorst, Christiana Albertina University of Kiel, Germany
Authors
Tiina Alamäe, University of Tartu, Estonia
Marta M. Andrade, Faculdade de Ciências da Universidade de Lisboa, Portugal
Ana Ardá, Centro de Investigaciones Biológicas, CSIC, Madrid, Spain
Juan M. Benito, Instituto de Investigaciones Quı´micas, CSIC - Universidad de
Sevilla, Spain
M. Álvaro Berbı́s, Centro de Investigaciones Biológicas, CSIC, Madrid, Spain
Nele Berghmans, Rega Institute for Medical Research, Belgium
Davide Bini, University of Milano-Bicocca, Milan, Italy
Pilar Blasco, Centro de Investigaciones Biológicas, CSIC, Madrid, Spain
Karin Bodewits, Ludwig-Maximilians-Universität, Munich, Germany
Paz Briones, CSIC, IBC, Sección de Errores Congénitos del Metabolismo,
Barcelona, Spain
Vasco Cachatra, University of Lisbon, Portugal
Valérie Calabro, Aix-Marseille University, France
Fernando Calais, Hospital São José, Lisboa, Portugal
Angeles Canales, Centro de Investigaciones Biológicas, CSIC, Madrid, Spain
F. Javier Cañada, Centro de Investigaciones Biológicas, CSIC, Madrid, Spain
Alice Capitoli, University of Milano-Bicocca, Milan, Italy
Mylène A. Carrascal, Universidade Nova de Lisboa, Portugal
Laura Cipolla, University of Milano-Bicocca, Milan, Italy
Fabio Dall’Olio, Università di Bologna, Italy
Anthony De Soyza, Newcastle University, UK
Flaviana Di Lorenzo, Università di Napoli Federico II, Italy
Sandrine Donadio-Andréi, Aix-Marseille University, France
Nassima El Maı̈, Aix-Marseille University, France
Amalia M. Estévez, CIQUS, University of Santiago de Compostela, Spain
Juan C. Estévez, CIQUS, University of Santiago de Compostela, Spain
Ramón J. Estévez, CIQUS, University of Santiago de Compostela, Spain
Ma Carmen Fernández-Alonso, Centro de Investigaciones Biológicas, CSIC,
Madrid, Spain
José G. Fernández-Bolaños, Universidad de Sevilla, Spain
Vanessa Ferreira, Portuguese Association for CDG and other Rare Metabolic
Diseases, Portugal
Luca Gabrielli, University of Milano-Bicocca, Milan, Italy
José M. Garcı́a Fernández, Instituto de Investigaciones Quı´micas,
CSIC - Universidad de Sevilla, Spain
Ana M. Gómez, IQOG-CSIC, Madrid, Spain
Alejandro González-Benjumea, Universidad de Sevilla, Spain
Chloé Iss, Aix-Marseille University, France
Jesús Jiménez-Barbero, Centro de Investigaciones Biológicas, CSIC, Madrid, Spain
Rosa Lanzetta, Università di Napoli Federico II, Italy
Sandra Li, Rega Institute for Medical Research, Belgium
J. Cristóbal López, IQOG-CSIC, Madrid, Spain
Óscar López, Universidad de Sevilla, Spain
Cristina Lupo, University of Milano-Bicocca, Milan, Italy
Andres Mäe, University of Tartu, Estonia
Filipa Marcelo, Centro de Investigaciones Biológicas, CSIC, Madrid, Spain
Karin Mardo, University of Tartu, Estonia
Sergio Martos, Universidad de Sevilla, Spain
Inés Maya, Universidad de Sevilla, Spain
Penélope Merino-Montiel, Universidad de Sevilla, Spain
Antonio Molinaro, Università di Napoli Federico II, Italy
Francesco Nicotra, University of Milano-Bicocca, Milan, Italy
Ana Oliete, Universidad de Sevilla, Spain
Ghislain Opdenakker, Rega Institute for Medical Research, Belgium
Carmen Ortiz Mellet, Universidad de Sevilla, Spain
Stefan Oscarson, University College Dublin, Ireland
José M. Otero, CIQUS, University of Santiago de Compostela, Spain
Amélia P. Rauter, University of Lisbon, Portugal
Catherine Ronin, Aix-Marseille University, France
Laura Russo, University of Milano-Bicocca, Milan, Italy
Paulo F. Severino, Universidade Nova de Lisboa, Portugal and Università di
Bologna, Italy
Alba Silipo, Università di Napoli Federico II, Italy
Mariana Silva, Universidade Nova de Lisboa, Portugal
Raquel G. Soengas, University of Aveiro, Portugal
Markus Sperandio, Ludwig-Maximilians-Universität, Munich, Germany
Francesca Taraballi, University of Milano-Bicocca, Milan, Italy
Clara Uriel, IQOG-CSIC, Madrid, Spain
Jo Van Damme, Rega Institute for Medical Research, Belgium
Paula A. Videira, Universidade Nova de Lisboa, Portugal
Maria-Antonia Vilaseca, Guia Metabólica, Esplugues de Llobregat, Spain
Triinu Visnapuu, University of Tartu, Estonia
Ulrika Westerlind, Leibniz Institute for Analytical Sciences, Dortmund, Germany
Alina D. Zamfir, National Institute for Research and Development in Electro-
chemistry and Condensed Matter, and Aurel Vlaicu University of Arad, Romania
If you buy this title on standing order, you will be given FREE access
to the chapters online. Please contact E-mail: sales@rsc.org with
proof of purchase to arrange access to be set up.
Thank you.
ISBN: 978-1-84973-439-4
ISSN: 0306-0713
DOI: 10.1039/9781849734769
A catalogue record for this book is available from the British Library
& The Royal Society of Chemistry 2012
All rights reserved
Apart from fair dealing for the purposes of research or private study for
non-commercial purposes, or for private study, criticism or review, as
permitted under the Copyright, Designs and Patents Act, 1988 and the
Copyright and Related Rights Regulations 2003, this publication may not be
reproduced, stored or transmitted, in any form or by any means, without the
prior permission in writing of The Royal Society of Chemistry, or in the case of
reproduction in accordance with the terms of the licences issued by the
Copyright Licensing Agency in the UK, or in accordance with the terms of the
licences issued by the appropriate Reproduction Rights Organization outside
the UK. Enquiries concerning reproduction outside the terms stated here should
be sent to The Royal Society of Chemistry at the address printed on this page.
Published by The Royal Society of Chemistry,

Thomas Graham House, Science Park, Milton Road,
Cambridge CB4 0WF, UK
Registered Charity Number 207890
For further information see our web site at www.rsc.org

Preface
DOI: 10.1039/9781849734769-FP005
Carbohydrate research plays a remarkable role in chemistry and biology

owing to the unique molecular features of the saccharides. They are mul-
tifunctional and stereochemically enriched molecules offering a superb
structural diversity to serve as molecular scaffolds or key intermediates in
the development of new drugs and novel carbohydrate-based or -functio-
nalized materials for a variety of applications. In addition, carbohydrates
are the molecular basis for a multitude of biological processes that occur in
states of health as well as disease. Understanding the chemistry and biology
of this class of molecules will facilitate new options for the treatment of
medical conditions for which no cure exists to date. Thus in this volume,
glycochemistry and glycobiology topics have been combined to document
the latest findings in carbohydrate research and demonstrate the contribu-
tions of organic chemistry, modern analytics, biological and biochemical
expertise to the increasingly important field of glycomics.
Firstly a modified polysaccharide, chlorite-oxidized oxyamylose, is
described as an immunomodulator with therapeutic implications for acute
and chronic inflammation, and also cancer. A chapter that focuses
on lipopolysaccharide in cystic fibrosis (CF)-related pathogens, namely
Pseudomonas aeruginosa and Burkholderia spp. reports on the structural
investigation of these biomolecules, their structural adaptation to the host
tissues as well as the importance of their structural features in the clinical
management of CF.
Development of carbohydrate vaccines against infections and cancer
continues to be a challenge for glycoexperts all over the world. Synthesis of
inner core lipopolysaccharides for vaccines against Gram-negative bacterial
infections and that of MUC1 glycopeptides conjugated to different immu-
nostimulants with promising results regarding the antibodies induced with
these synthetic vaccines against cancer will be highlighted here.
The immune response to invading pathogens during inflammation is
crucial in cell biology. Thus, the role of carbohydrate decorations in
leukocyte recruitment will be reviewed, putting a strong focus on post-
translational modification by sialic acids. New findings emphasizing the
influence of carbohydrate decoration on the regulation of inflammatory
response will be discussed. New interesting therapeutic approaches in the
treatment of acute and chronic inflammatory diseases are being offered.
Recent progress on glycoengineering, an advanced technology based on a
glycosylation strategy to optimize protein drugs, is reported. Impact of
N-glycosylation on therapeutic proteins, namely that of glycans on drug
bioavailability, glycoprotein biopotency, drug immunogenicity, protein
folding and epitope expression as well as production, safety and regulatory
aspects are discussed. Major advances in glycoengineering are reported with
an emphasis on N-glycosylation control to identify the conditions that
promote an optimal glycoform profile and reproducibility of glycomodifi-
cation. In addition, identification and cloning of glycosyltransferase genes
Carbohydr. Chem., 2012, 38, vii–viii | vii

c The Royal Society of Chemistry 2012
are described as new tools for manipulation of expression systems in order
to further improve the glycoform profile, in particular engineering of core
fucosylation and sialylation.
Glycosylation has significant implications in health and disease. Hence,
one chapter is dedicated to congenital disorders of glycosylation, which are
a group of disorders of abnormal glycosylation of N-linked oligosacchar-
ides caused by deficiency in 29 different enzymes in the N-linked oligo-
saccharide synthetic pathway. The relevance and key aspects of the
glycosylation changes associated with bladder cancer are highlighted in
another chapter of this volume.
Interdisciplinarity of the glycosciences is also demonstrated by a con-
tribution on novel approaches for the production of levansucrases, bacterial
extracellular enzymes that act on sucrose producing b-2,6-linked fructans.
Biochemical characterization of this protein encoded in the genome from
Pseudomonas bacteria, and an innovative mass spectrometric study of the
reaction products permits to identify the produced potentially prebiotic
fructooligosaccharides from sucrose or raffinose.
NMR is currently a potent methodology to analyse sugar conformation
and to study binding interactions. New advances from the NMR metho-
dological viewpoint for analysis of saccharide conformation are described in
a chapter, in which examples are given for oligo- and polysaccharides,
glycopeptides, glycomimetics, and also carbohydrate-protein, carbohy-
drate-carbohydrate, and carbohydrate-nucleic acid interactions.
The contribution of glycochemistry to innovation in glycosciences is
shown in chapters 10–17. Here imino sugar glycosidase inhibitors, carba-
sugars, multivalent glycoconjugates, including glycodendrimers, glycona-
notubes, and glyconanoparticles, and their uses in medicinal chemistry, as
well as artificial saccharide-based and saccharide-functionalized gene
delivery systems are presented.
Highly functionalized exo-glycals used for the generation of molecular
diversity in a chemoselective manner, namely for the preparation of fur-
anose-based libraries with three or more sites for molecular diversity are
discussed. A chapter on siderophores that are based on monosaccharides
that have proven effective for Gram-negative bacteria and mycobacteria,
and the chapter on biomaterials, in particular on the so-called smart
materials, that can modulate and control cell behaviour, complete the
volume.
Volume 38 of Carbohydrate Chemistry – Chemical and Biological
Approaches contains contributions ranging from glycochemistry to glyco-
biology. They have been authored mostly by scientists that are members of
the European Science Foundation Network Euroglycoforum. This collec-
tion demonstrates in a meaningful way how the interdisciplinary approach
of an international glyconetwork can advance the field of carbohydrate
research in Europe and worldwide.
We hope you will enjoy the beauty of the ‘‘sweet’’ chemistry and biology
presented herein!
Ame´lia Pilar Rauter and
Thisbe K. Lindhorst
viii | Carbohydr. Chem., 2012, 38, vii–viii

CONTENTS
Cover
Tetrahydropyran-enclosed
ball-and-stick depiction of a glucose
molecule, and (in the background)
part of an a-glycosyl-(1-4)-D-glucose
oligosaccharide and a glycosidase,
all representative of the topics
covered in Carbohydrate Chemistry –
Chemical and Biological Approaches.
Cover prepared by R. G. dos Santos.
Preface vii
Ame´lia Pilar Rauter and Thisbe K. Lindhorst
Applications of glycobiology: biological and immunological effects of a 1

chemically modified amylose-derivative
Ghislain Opdenakker, Sandra Li, Nele Berghmans and Jo Van Damme
1 Introduction 1
2 Historical breakthroughs and examples 2
3 Polysaccharides and derivatives 5
4 An historical finding in virology? 5
5 COAM does not induce interferon 6
6 COAM is an immunomodulator 7
7 Therapeutic implications for acute and chronic 8
inflammation and cancer
8 Conclusions and future perspectives 9
Abbreviations 10
Acknowledgements 10
References 10
Lipopolysaccharide structure and biological activity from the cystic 13

fibrosis pathogens Burkholderia cepacia complex
Anthony De Soyza, Flaviana Di Lorenzo, Alba Silipo, Rosa Lanzetta
and Antonio Molinaro
1 Introduction 13
Carbohydr. Chem., 2012, 38, ix–xiv | ix

2 Basic Lipopolysaccharide structure 16
3 The detailed chemical structure of the P. aeruginosa 20
LPS
4 Burkholderia cepacia complex and Cystic Fibrosis 23
5 General structural features of the core region of BCC 28
LPS
6 O-chain of BCC LPS 30
7 Conclusion 33
Acknowledegments 34
References 34
Synthesis of bacterial carbohydrate surface structures containing Kdo 40

and glycero-D-manno-heptose linkages
Stefan Oscarson
1 Introduction 40
2 Kdo-containing structures and Kdo donors 40
3 Glycero-D-manno-heptose-containing structures 48
References 59
Synthetic glycopeptides in vaccine development and antibody epitope 61

mapping
Ulrika Westerlind
1 Introduction 61
2 Protein conjugate vaccines 62
3 Build-in adjuvant vaccines 65
4 Dendrimer vaccines 68
5 Antibody epitope mapping 69
6 Conclusions 72
References 73
Posttranslational sialylation and its impact on leukocyte recruitment 75

during inflammation
Karin Bodewits and Markus Sperandio
1 Leukocyte recruitment cascade 75
2 Posttranslational modifications 78
3 Sialic acids and sialyltransferases (ST) 82
4 Sialylation-dependent functions of signalling and adhesion 88
relevant molecules
5 Conclusion and outlook 91
References 91
x | Carbohydr. Chem., 2012, 38, ix–xiv

Glycoengineering of protein-based therapeutics 94
Sandrine Donadio-Andreí, Chloe´ Iss, Nassima El Maı¨,
Vale´rie Calabro and Catherine Ronin
Introduction 94
1 Impact of N-glycosylation on therapeutic proteins 95
2 Glycosylation optimization 105
3 Conclusions 113
References 115
Congenital Disorders of Glycosylation (CDG): from glycoproteins 124

to patient care
Vanessa Ferreira, Paz Briones and Maria-Antonia Vilaseca
1 Introduction 124
2 Basic principles of protein glycosylation 124
3 Congenital disorders of glycosylation (CDG) 128
4 Concluding remarks and future perspectives: CDG as a 144
challenge for glycobiomedicine in the next decade!
5 Highlights 145
6 Further information 146
Abbreviations 146
Acknowledgements 147
References 147
Bladder cancer–glycosylation insights 156

Paulo F. Severino, Mariana Silva, Myle`ne A. Carrascal,
Fernando Calais, Fabio Dall’Olio and Paula A. Videira
1 Introduction 156
2 Bladder cancer: epidemiology and current treatment 157
3 General aspects of glycosylation 158
4 Final considerations/conclusions 168
Abbreviations 169
References 169
Levansucrases of Pseudomonas bacteria: novel approaches for protein 176

expression, assay of enzymes, fructooligosaccharides and
heterooligofructans
Tiina Alamäe, Triinu Visnapuu, Karin Mardo, Andres Mäe and
Alina D. Zamfir
1 Levansucrase genes and proteins 176
2 Heterologous expression of levansucrases from Pseudomonas 177
syringae pv. tomato with two different expression systems
Carbohydr. Chem., 2012, 38, ix–xiv | xi

3 Biochemical properties of the levansucrases: substrate 178
specificity and kinetic parameters
4 Predicted catalytic triad residues and 3D structures of Lsc1, 179
Lsc2, Lsc3 and LscA proteins
5 Lsc2, Lsc3 and LscA produce not only levan, but also 180
FOS that can be detected by a novel mass spectrometric
method
6 Isolation and novel screening methods for levansucrase 186
mutants
7 Concluding remarks and future perspectives 189
Acknowledegements 190
References 190
Recent advances on the application of NMR methods to study the 192

conformation and recognition properties of carbohydrates
Ana Ardá, M. Álvaro Berbı´s, Pilar Blasco, Angeles Canales,
F. Javier Cañada, Ma Carmen Fernández-Alonso, Filipa Marcelo and
Jesús Jimeńez-Barbero
1 Introduction 192
2 The access to new NMR parameters and methodological 192
developments
3 Applications. Saccharides in solution 194
4 Applications. The interaction of saccharides with other 202
natural and synthetic molecules
Acknowledgments 210
References 210
Glycosidase inhibitors: versatile tools in glycobiology 215

Óscar López, Pene´lope Merino-Montiel, Sergio Martos and
Alejandro González-Benjumea
1 Introduction 215
2 Polyhydroxylated pyrrolidines 216
3 Six-membered ring-based imino sugars 219
4 Azepanes and azetidines 226
5 Bicyclic imino sugars 231
6 Thio, seleno and carbasugars as glycosidase 245
inhibitors
7 Miscellaneous 250
8 Pharmacological activities of imino sugars 251
9 Concluding remarks 256
References 256
xii | Carbohydr. Chem., 2012, 38, ix–xiv

An overview of key routes for the transformation of sugars into 263
carbasugars and related compounds
Raquel G. Soengas, Jose´ M. Otero, Amalia M. Este´vez,
Ame´lia P. Rauter, Vasco Cachatra, Juan C. Este´vez and
Ramón J. Este´vez
1 Introduction 263
2 Strategies for the synthesis of carbasugars and 265
pseudo-carbasugars
References 297
Multivalent glycoconjugates in medicinal chemistry 303

Jose´ G. Fernández-Bolaños, Ine´s Maya and Ana Oliete
1 Introduction 303
2 Glycoclusters with a flexible core 303
3 Glycoclusters with a rigid core 308
4 Multivalent glycopeptides in immunotherapy 313
5 Glycoconjugates on tubular scaffolds 318
6 Glycodendrimer chips 318
7 Glyconanoparticles 320
8 Conclusions 331
References 331
Glycotransporters for gene delivery 338

Carmen Ortiz Mellet, Jose´ M. Garcıá Fernández and Juan M. Benito
1 Introduction 338
2 Carbohydrate-grafted cationic polymers and lipids 341
3 De novo designed carbohydrate-based polymers 346
4 Preorganized glycomaterials for gene delivery 354
5 Outlook and perspectives 368
Acknowledgments 368
References 369
Furanose-based templates in the chemoselective generation of 376

molecular diversity
Ana M. Gómez, Clara Uriel and J. Cristóbal López
1 Introduction 376
2 Design and synthesis of functionalized furanose-based 377
precursors
Carbohydr. Chem., 2012, 38, ix–xiv | xiii

3 Reactivity of functionalized furanose-based precursor 4a 380
4 Reactivity of the alkenyl halide moiety in functionalized 387
furanose-based precursors 4b,c
5 Three components assembly of functionalized furanose 388
derivatives 4b,c
6 Generation of furanosidic libraries possessing three – or 391
more – sites for molecular diversity
7 Conclusions 392
Abbreviations 393
References 393
Synthesis of carbohydrate-based artificial siderophores and their 398

biological applications
Marta M. Andrade and Ame´lia P. Rauter
1 Introduction 398
2 Artificial siderophores - background 400
3 Synthesis of Carbohydrate-based artificial siderophores 402
4 Study of siderophore activity 407
5 Conclusions and perspectives 412
References 414
Smart biomaterials: the contribution of glycoscience 416

Laura Cipolla, Laura Russo, Francesca Taraballi, Cristina Lupo,
Davide Bini, Luca Gabrielli, Alice Capitoli and Francesco Nicotra
1 Introduction: biomaterials and tissue engineering 416
2 Glycoscience for biomaterials design 424
Conclusions 440
References 440
xiv | Carbohydr. Chem., 2012, 38, ix–xiv

Applications of glycobiology: biological and
immunological effects of a chemically
modified amylose-derivative
Ghislain Opdenakker,* Sandra Li, Nele Berghmans
and Jo Van Damme
DOI: 10.1039/9781849734769-00001
Carbohydrate chemistry, oligosaccharide sequencing, synthesis technolo-

gies and glyco-engineering have helped to establish glycobiology alongside
molecular biology. However, examples of therapeutic implications of gly-
cobiology are limited to oligosaccharides. These include engineered anti-
bodies containing sialic acids, inhibition of leukocyte rolling effect in the
interactions between mucins and selectins and antiviral imino sugars. The
best example of a successful glycodrug is oseltamivir, Tamiflus, as an
inhibitor of the influenza virus neuraminidase. Polysaccharides, although at
first sight structurally less complex and biologically less challenging, are
interesting molecules with unexplored possibilities in biology and medicine.
Polysaccharides may form protease-resistant scaffolds for tissue engineering
and chemically modified polysaccharides (CMPs) have potent immuno-
modulating activities. This is illustrated here with chlorite-oxidized oxya-
mylose (COAM), a broad-spectrum antiviral agent. COAM interacts with
the chemokine system and thus can be used to modulate leukocyte com-
partmentalization in vivo. This knowledge has been used to alter the clinical
course of acute viral infections, cancer and experimental autoimmune
encephalomyelitis (EAE), an animal model of multiple sclerosis. The latter
example is proof-of-concept that CMPs constitute important probes to
study immune functions and are novel drugs applicable for specific disease
entities.
1 Introduction
From semantic viewpoint the term molecular biology is a pleonastic
expression and does not correspond with the present day interpretations.
Indeed, all biological processes are driven by molecules. What is meant by
molecular biology are processes driven by nucleic acids. Thus, molecular
biologists are stricto sensu researchers using DNA and RNA tools for
sequencing and synthesis, for practical uses in genetic engineering and for
therapy with recombinant drugs. Glycobiology (the biology of glycans) is a
more strict term in that it defines, by its name, the molecules under study,
namely carbohydrates. Why does the field of glycobiology then still need
promotion? Why do the DNA/RNA biologists generate more appreciation
and the glycobiologists less recognition? This is certainly not because DNA
and RNA are sugar-phosphate polymers. Maybe it is because of discovery
of the beauty of the linear connection between codons and amino acids.
Rega Institute for Medical Research, Minderbroedersstraat 10, B-3000 Leuven, Belgium.
E-mail: ghislain.opdenakker@rega.kuleuven.be
Carbohydr. Chem., 2012, 38, 1–12 | 1

Maybe it is also because functional aspects of a branched glycan tree are not
yet understood. As long as the interconnections between RNA/DNA
biology with glycobiology are not yet hold by a solid stem, scientists need to
invest time and resources in finding important examples in glycobiology.
This is simply to state that the real basis of glycobiology, for instance
relating structures to functions, is a barrier for many scientists and needs
further promotion. The original concepts that glycoforms and glycotypes
confer different functions to proteins must be seen as eye-openers [1–2].
These concepts, well elaborated for immunoglobulin-G (IgG) and tissue-
type plasminogen activator (t-PA), are becoming apparent for many other
glycoproteins. Aside effects of glycosylation on the specific activity of t-PA
and functional aspects of the glycans on the Fc part of IgG (vide infra),
glycoprotein folding, trafficking, recognition and involvement in the
immune system demonstrate that oligosaccharides are endowed with mul-
tiple functions [3–6].
The successes in technology progresses in nucleic acid and protein
sequencing and synthesis are evident and their impacts on medicine,
industry and society are obvious. The field of RNA/DNA biology has the
intrinsic advantage of a limited number of building blocks, with limited
natural chemical modifications of the five bases. Protein biology, based on
about 20 building blocks, is already more complex and subject to many
more natural chemical modifications: proteolytic processing, phosphoryla-
tion and dephosphorylation, citrullination, isoprenylation, acetylation/
deacetylation, to name a few. Glycosylation is nowadays recognized as the
most diverse and complex posttranslational modification of proteins.
Glycobiology forms the next, and perhaps the most challenging level.
Intrinsically, carbohydrate biochemistry is also based on a limited number
of natural monosaccharides, but their chemical modifications are multifold.
In addition, the glycosidic bonding of monosaccharides is subject to various
linkages and the involved carbon atoms form stereochemical centers. The
branching of oligo- and polysaccharides into arborized structures often
supersedes our imagination capacity, which is too much trained for linear
DNA, RNA and protein sequences. However, in practice in mammalian
systems a limited set of monosaccharides are used. Even so the linkage
variation creates an extraordinary diversity of structures.
2 Historical breakthroughs and examples

If glycobiology is to succeed as the science of the 21st century, then one
needs to define new breakthroughs. Happily one can walk in the footsteps of
those who prepared and enhanced the field with exoglycosidase oligo-
saccharide sequencing [1, 2], with mass spectrometry [7] and NMR analysis
of sugars [8], with lectin studies [9] and with glyco-engineering. The dis-
covery of congenital disorders of glycosylation [10] not only points to future
therapeutic applications, but also has yielded a bridging function between
glycobiology and nucleic acid biology.
Oligosaccharide sequencing evolved over the last 30 years from an
extremely labor-intensive discipline towards an efficient analytical method
in the hands of experts. Pioneering work by the groups of Akira Kobata
2 | Carbohydr. Chem., 2012, 38, 1–12

(University of Tokyo) [11] and of Raymond Dwek (Oxford Glycobiology
Institute) [12] led to the understanding of the proinflammatory effects of
agalactosyl IgGs in rheumatoid arthritis [2]. This knowledge was com-
plemented with the findings of anti-inflammatory sialic acids in IgGs [13].
Together, these breakthroughs can be summarized in one sentence. Oligo-
saccharides attached to proteins fine-tune the biological activities of these
molecules. This should be not compared with an on/off switch (as is often
the case with protein phosphorylation), but rather with the selection of a
specific program or tuning. Presently, oligosaccharide exoglycosidase
sequencing has been perfected to such a level [14, 15] that broad-spectrum
and high-throughput analysis and glycomics are within reach. One next
breakthrough will come from coupling of this type of sophisticated analysis
with proteomics, genomics and transcriptomics.
A second aspect of structural analysis is mass spectrometry and NMR
analysis. Whereas mass spectrometry per se is nowadays routinely used in
proteomic analysis [16], for oligosaccharide analysis the identification of
structural details by mass spectrometry remained more complicated for the
simple reason that, for example, the various hexoses maintain the same
molecular mass. Nevertheless, sophisticated mass spectrometry technology
for sugars has been developed, is presently the expertise of few specialists
[7, 17], but is becoming common practice, because the combination of
high-performance liquid chromatography and mass spectrometry of oligo-
saccharides is very powerful and rapid. Considerable growth is expected in
this area and basic mass spectrometry analysis is already becoming classical
technology for structural analysis. Unfortunately and as is the case for other
applications of mass spectrometry, high-end equipment is rather expensive
and a limited number of specialized centers, homing dedicated specialists
and equipped with such instruments, can answer easily any exotic demands
for the analysis of complex glycoconjugates. Similar economic considera-
tions may apply for NMR analysis, with the additional practical point
that the amounts of purified glycoconjugates needed for NMR analysis
supersede considerably those for mass spectrometry identification.
A third aspect is related to practical aspects, which we best place under
the term glyco-engineering. Agalactosyl [2] and sialyl IgGs [13] have already
been mentioned as disease promoting and disease-limiting factors, res-
pectively, in inflammation. Such knowledge is successfully addressed by
glyco-engineering companies making optimized recombinant glycoprotein
preparations [18]. Another example is related to the classical ABO blood
group system. Nobel Prize laureate Karl Landsteiner was a glycobiologist
avant-la-lettre. Present-day glyco-engineers have managed to modify the
ABO blood group antigens [19], heralding novel possibilities in transfusion
medicine. Because the immunological insights with blood transfusions
formed the basis of organ transplantation medicine, it is not excluded that
the glyco-engineering of the ABO blood group antigens forms the basis of
new applications in xenotransplantation, where galactosylation is a major
hinder in the application of pig donor organs [20].
A fourth aspect to be recognized as historical in glycobiology is the study
of lectins as probes for sugars. Plant lectins, recognizing in defined ways
oligosaccharides, have been widely used in medical diagnostics, often before
Carbohydr. Chem., 2012, 38, 1–12 | 3

their mechanism of action was known [21, 22]. Lymphoblast transformation
with phytohaemagglutinin for karyotyping purposes and definition of
chromosomal aberrations was used long before the cytokines that are
induced in this process were identified as interleukin-2, interleukin-6 and
others. Lectins are also used for disease diagnosis in histochemical analysis
of cancer and, more recently, with the use of lectin microarrays [23]. The
literature on this topic is vast, exemplary reviews are available for detailed
and general information [24, 25]. In contrast, the information about lectins
as therapeutics is still rather sketchy. After the discovery of selectins, the
mammalian lectins that regulate rolling of leukocytes at an early step in the
inflammatory response, selectin knockout mice were developed to study
immunophenotypes (how these mice differ from wildtype animals with
reference to immunological parameters) and attempts were made to develop
novel anti-inflammatory agents on the basis of these lectin-glycan interac-
tions [26–28]. A second example about glycotherapy relates to viral infec-
tions. Human immunodeficiency virus is enveloped in a glycocoat, limiting
access to protein (peptide) epitopes. One way to flag this virus is with
mannose-recognizing lectins, a strategy that is investigated with increasing
intensity [29].
An important way to enhance the field of glycobiology is to define
new medical applications. These may involve deficiencies, resulting from
aberrant glycosylation, in need for substitution therapy or storage diseases
with a pathogenic role of accumulated saccharides, in which reduction
forms a treatment strategy. An important class of rare diseases is the family
of congenital diseases of glycosylation (CDG) [10]. However, a critical
aspect to recognize is the fact that CDGs create the link between nucleic
acid and protein biology and glycobiology. Identified CDGs, resulting
from the mathematics of evolution (combinations of random mutations
that result in life offspring with clinical phenotypes), demonstrate that
glycosylation is a biological process of vital importance. In addition, the
creation of knockout mouse models of CDGs is an important step to
understand better the biological importance of sugars, but may also form a
solid basis to define treatments of these rare diseases, both in cases of
deficiencies [30] and of substrate reduction therapy in storage diseases, like
Gaucher disease [31].
Finally, we come to the level of commonly used drugs. One viral target
enzyme stands out as a glycobiological example: the influenza virus neur-
aminidase with its inhibitors oseltamivir (Tamiflus) and zanamivir
(Relenzas). The concept is simple, but the efforts done are considerable:
definition of a microbial target enzyme (sufficiently different from sister
enzymes in the host) and screening for inhibitors [32]. It does not take much
imagination that future antivirals and other antimicrobial agents may be
developed on the basis of other microbial sugar modifying enzymes. In
virology, such examples of interesting glycotarget enzymes may be restricted
because of the limitation of the coding capacity of viral genomes and,
therefore, the influenza neuraminidase will keep the lead for a while.
However, in the world of bacteria, fungi and parasites an enormous
potential exists to discover target enzymes, generate inhibitors and define
new medicines.
4 | Carbohydr. Chem., 2012, 38, 1–12

Most of the examples mentioned so far relate to oligosaccharides. What is
the status about polysaccharides?
3 Polysaccharides and derivatives

Although changes are emerging with the recognition of the importance of
polysaccharide vaccines, in biomedical education the time spent to study
polysaccharides is inversely correlated with their abundance in nature.
Cellulose [polyb(1-4)D-glucose], amylose [polya(1-4)D-glucose] and bran-
ched amylopectin in starch form most biomass and energy supply for most
organisms. Intestinal resorption of starch glucose (from amylose and
amylopectin) as monosaccharides and disaccharides and conversion into
glycogen is basic medical knowledge that can be linked to rare glycogen
storage diseases and to metabolic disease number one, i.e. diabetes. Defi-
ciencies of amylase and disaccharidases lead to alterations of the gut
microbiomes and to gastrointestinal disorders as is the case with specific
polysaccharide derivatives. Indeed, dextran sodium sulphate (DSS) is
commonly used to induce inflammatory bowel disease in mice (used as an
animal model for Crohn’s disease or ulcerative colitis). In other words,
polysaccharide derivatives may be harmful or not, depending on their
degradability and their location.
4 An historical finding in virology?

After the discovery of interferon (IFN) in 1957, great hope was generated
for its use as therapy for viral diseases and mass production lines were
explored ending in the recombinant expression of IFN and use in the clinic
[33]. IFNs were not developed into first choice antivirals. Eventually, IFN-a
from leukocytes and IFN-b from fibroblasts became first choice anti-
inflammatory drugs for the treatment of multiple sclerosis. Recombinant
IFN-g, also dubbed immune interferon, because it is produced by lym-
phocytes and natural killer cells, found its way to the treatment of chronic
granulomatous disease [34].
Piet De Somer (1917–1985), founder of the Rega Institute (Fig. 1) was an
insightful man because he thought about and tried to solve the IFN pro-
duction problem (before its cloning and expression [35]) by finding ways to
induce endogenous IFN. He admired and followed his scientist example,
Karl Landsteiner, by trying to use simple but practical rules. Karl
Landsteiner, who discovered the ABO blood group system, used simple
serological tests for discovery.
The aftermath of the second world war was characterized by discoveries
of many new mold-derived antibiotics, when antiviral IFN was discovered
upon its inducibility by viruses. IFNs protect neighbouring cells by inducing
an antiviral state and the virologists in the early 1960s started to search for
inducers that could replace viruses. In those booming days of DNA and
RNA research, screening of mold or bacterial extracts and synthetic
molecules for antiviral activity was a commonly used technique. Penicillium
products and synthetic polyanionic polymers were found to possess broad
spectrum antiviral activity and this incited De Somer to search for and test
other polyanions for antiviral activity. De Somer explored synthetic
Carbohydr. Chem., 2012, 38, 1–12 | 5

Fig. 1 Rector Piet De Somer (1917–1985) and the Rega Institute for Medical Research,
founded in 1954. Drawings are by Gerard Thijs, graphic artist, Leuven.
polyanions, such as polymethacrylates and polyacrylates, as antiviral agents

[36] and as inducers of interferon with some success [37]. Later it was found
that nucleic acids, such as synthetic double stranded RNA, induce the
production of IFN (most probably by mimicking the replicative inter-
mediate forms of RNA viruses). Double-stranded RNA is a polyanionic
molecule and like polyriboinosinic acid:polyribocytidylic acid (poly IC),
synthetic polyanions were found to be toxic, presumably by their non-
degradability. De Somer then gave the chemists of the Rega Institute the
task to manufacture a biodegradable polyanionic molecule that could be
used as an antiviral agent and/or interferon inducer: chlorite-oxidized
oxyamylose or COAM was born [38]. The chemists used amylose as starting
material in a stepwise oxidation reaction with openings of the glucose rings
(Fig. 2) to yield the polycarboxylic acid structure, now known as COAM.
Quickly, the product was tested against viruses and found to be a potent
and broad-spectrum antiviral agent [38–48]. In those days, this was often
not done with cell-based assays and permissive viruses, but instead by in vivo
testing (Table 1). COAM was supposed to induce IFN, although from the
time of the first studies onwards it was not clarified that the observed serum
IFN levels in mice treated with COAM and infected with virus were indeed
derived from induction by either the virus or by COAM. Because of some
inconclusive discrepancies in the COAM literature [39, 49], the question
whether COAM itself induces IFN or enhances IFN production by other
inducers remained open. In fact, data obtained already in 1976 from in vivo
experiments, corroborating earlier studies with COAM [42], remained
hidden in our laboratory notebooks till publication in 2010 [50].
5 COAM does not induce interferon

About 35 years after its synthesis [38], we reiterated the COAM project in
2006 using the now available powerful techniques of genetic engineering.
First, we applied a simple in vitro virus infection model with the picorna-
virus mengovirus on mouse fibroblastoid L929 cells. COAM protected
6 | Carbohydr. Chem., 2012, 38, 1–12

Fig. 2 Chemical synthesis of COAM from amylose. The steps of glucose ring opening by
periodate and further oxidation with chlorite into a polycarboxylate structure is illustrated
(Adapted from P. Claes et al., 1970, reference 38).
dose-dependently the cell cultures against mengovirus, but only poly IC

induced an antiviral state by induction of IFN-b. We did not observe
induction of IFN bioactivity or of IFN mRNAs by COAM. By performing
time-of-addition experiments with COAM, we deduced that COAM
acted at the stage of virus entry. It was also noticed that rather high
concentrations (1 mg/ml) of COAM were needed to exert an in vitro
antiviral effect [51].
6 COAM is an immunomodulator
The in vivo efficacy of COAM against many viruses (Table 1) was much
better than what was expected on the basis of the in vitro antiviral profile.
For those reasons immune cells and molecules were thought to be involved
and were studied. The conclusions that were deduced from these studies
included (i) COAM acts mainly on myeloid cells (neutrophils and macro-
phages) rather than on lymphocytes and (ii) intraperitoneal injection of
COAM leads to massive neutrophil influx, hinting to an interaction with
neutrophil chemokines. Since the most potent neutrophil chemokine in the
mouse is granulocyte chemotactic protein-2 (GCP-2) [52], the induction and
binding of GCP-2 by COAM was investigated and the link with chemokines
was established. It was concluded that COAM induces and binds GCP-2
and thus establishes a potent chemokine gradient for myeloid cells to the
peritoneal cavity where the local virus infection is contained [50]. The
information about a CMP drug that protects against lethal virus infections
after prophylactic parenteral administration is perhaps not so important
for common acute virus infections. However, because rescue (more than 90
Carbohydr. Chem., 2012, 38, 1–12 | 7

Table 1 Antiviral effects of COAM
Dosis/
Virusa Species (treatment)b concentrationc Effectd Reference
Coxsackie B4 Mice (profyl. 24 h) 2 mg/mouse ip. Protection 53

Foot-and-mouth ip. Mice (profyl. 24 h) 5–60 mg/kg ip. Protection 40
Foot-and-mouth ip. Swine (profyl. 24 h) 120 mg/kg ip. No effect 40
Friend leukemia Mice (profyl.) 0.5 mg/mouse Improvement 46
Hog cholera Swine (profyl. 24 h) 120 mg/kg ip. Protection 40
HSV-1 in eye Rabbit (profyl. 4 h) 50 mg/ml topic. Protection 40
HSV-1 in. Mice (profyl. 24 h) 0.8–4 mg/mouse ip. Protection 47
Influenza in. Mice (profyl. 24 h) 2 mg/mouse ip. Protection 39
Influenza ip. Mice (profyl. 24 h) 2 mg/mouse ip. Protection 41
Mengo ip. Mice (profyl. 24 h) 2 mg/mouse ip. Protection 40
Mengo ip. Mice (profyl. W16 h) 2 mg/mouse ip. Protection 42
Mengo ip. Mice (profyl. 24 h) 2 mg/mouse ip. Protection 50
Mengo L929 cells (in vitro) 1 mg/ml CPE inhibition 51
Moloney mouse Mice (profyl. 24 h) 0.5 mg/mouse ip. Protection 43
sarcoma im.
Newcastle disease iv. Mice (profyl.) 1–10 mg/mouse ip. Protection 44
PVX leaf discs Plant (posti.) 1% solution No effect 48
Rabies im. Mice (3 h posti.) 0.3 mg/mouse Worsening 45
Semliki Forrest ip. Mice (profyl. 24 h) 2 mg/mouse ip. Protection 39
TMV leaf discs Plant (posti.) 1% solution Protection 48
Vaccinia iv. Mice (profyl. 24 h) 2 mg/mouse ip. Protection 38
Vaccinia iv. Mice (profyl. 24 h) W55 mg/mouse iv. Protection 39
Vesicular stomatitis MEF (in vitro) 10 mg/ml CPE inhibition 38
Vesicular stomatitis Mice (profyl.) 10 mg/mouse No effect 44
in.
a
HSV-1: Herpes simplex virus type 1; TMV: Tobacco mosaic virus; PVX: Potato virus X.
b
profyl.: prophylactic, before virus; posti.: postinfection.
c
iv.: intravenous; ip.: intraperitoneal; in.: intranasal; im.: intramuscular.
d
CPE: cytopathic effect.
percent survival without disease sequels) was observed after lethal virus
challenge by pretreatment with a purified COAM preparation, the CMP
preventive action may be crucial to contain clusters of lethal virus infec-
tions, including those with Ebola and other deadly viruses [50]. In another
model of Coxsackievirus B4 pancreatic infection it was corroborated that
COAM has potent antiviral effects, most probably mediated by myeloid
inflammatory cells [53].
7 Therapeutic implications for acute and chronic inflammation and cancer

On the basis of the mentioned experimental data, the question was asked
whether the polysaccharide derivative COAM may be applicable in other
diseases. De Somer’s wish of using IFN-b as a potent antiviral agent did not
really materialize. Instead, IFN landed as a real drug for the treatment of
multiple sclerosis [33].
At the time when it was still controversial whether COAM induces IFN
and not yet clear that COAM acted in a different way, we investigated
whether COAM might protect against experimental autoimmune ence-
phalomyelitis (EAE), an animal model of multiple sclerosis. It is worth
8 | Carbohydr. Chem., 2012, 38, 1–12

mentioning that these experiments were also intended to investigate the
clinical effects of COAM in virus-free conditions.
The major conclusions from studies with a hyperacute model of EAE
formed the basis of a doctoral dissertation (Dr. N. Berghmans, 2008,
University of Leuven, Belgium, promoter Prof. H. Heremans). COAM did
not induce IFN-a, IFN-b or IFN-g in vivo, but possessed a significant
protective effect against hyperacute EAE. Intraperitoneal injection of
COAM was locally chemotactic for leukocytes with such potency that it
diminished the leukocyte numbers significantly in the central nervous sys-
tem (brain and spinal cord). Because of these observations, we proposed a
new immunotherapeutic strategy to treat acute and chronic CNS inflam-
mation and called this ‘‘cell deviation therapy with an immunostimulating
glycomimetic’’. This research is presently continued and firm evidence has
been established that COAM also protects against autoimmune EAE,
induced by myelin oligodendrocyte glycoprotein-derived peptide in
EAE-susceptible IFN-g knockout mice [54] and [Berghmans et al., in
preparation].
Another type of application was investigated in tumor biology. As
already described in Table 1, COAM protected against Moloney mouse
sarcoma virus infections [43]. This was interpreted as an antiretroviral
effect, and also as an antitumoral effect, because both tumor incidence and
lethality caused by tumor growth were reduced by COAM. Moreover, by
now it is clear that inflammation is a critical component in tumor biology
[55] and that myeloid cells, alongside lymphocytes, are key players in tumor
biology [56]. For these reasons, we took the simple approach of testing
COAM against melanoma in a syngeneic mouse model. It was found that
COAM reduced significantly the tumor burden of dermal melanomas when
locally injected early in tumor development [57].
8 Conclusions and future perspectives

In comparison with many studies on the biology of complex oligosaccharide
structures, polysaccharides and derivatives have not yet attracted much
interest from industry. Nevertheless several interesting principles may be
demonstrated with the example of the amylose-derivative COAM. First, old
literature and hypotheses (about polysaccharides and derivatives) should be
tested with all novel technologies at our disposal. For instance, structural
analysis can now be executed with the tools of exoglycosidase sequencing
and mass spectrometry and biological analyses may rely on the use of RT-
PCR, transcriptional micro-arrays and in vivo animal models. Second, on
the basis of biological discoveries new therapies may be invented. For
instance, on the basis of the discovery that COAM (induces and) binds
chemokines and thus may create unseen potent endogenous chemokine
gradients, we discovered that it is possible to call leukocytes out of the CNS
back into the circulation and the (immunological) periphery to cure mouse
EAE. Such proof of principle may pave the way to new treatments of acute
and chronic neuroinflammations and should definitely stimulate basic and
applied research on polysaccharides.
Carbohydr. Chem., 2012, 38, 1–12 | 9

Abbreviations
CMP chemically modified polysaccharide
COAM chlorite-oxidized oxyamylose
CDG congenital disorder of glycosylation
CPE cytopathic effect
EAE experimental autoimmune encephalomyelitis
HSV-1 herpes simplex virus type 1
IgG immunoglobulin-G
IFN- interferon-
im. intramuscular
in. intranasal
ip. intraperitoneal
iv. intravenous
poly IC polyriboinosinic acid:polyribocytidylic acid
t-PA tissue-type plasminogen activator
Acknowledgements
This overview is a tribute to the late Professor Piet De Somer, founder of the
Rega Institute for Medical Research and an eminent scientist-entrepreneur
and university rector in Belgium. To commemorate his passing away, 25
years ago, the University of Leuven honored De Somer with a symposium
and the City of Leuven renamed its central place into ‘‘Piet De Somer
Square’’. The authors thank Professor Alfons Billiau (Leuven) and
Professor Raymond A. Dwek (Oxford) for inspiration and critically reading
and commenting on the manuscript and Professor Hubertine Heremans
(Leuven) for expertise in EAE and promotion of the COAM-project. This
work was funded by the ‘‘Geconcerteerde OnderzoeksActies’’ (GOA2012/
17) and by the Fund for Scientific Research of Flanders (FWO-Vlaanderen).
GO is a Steering Committee Member of the European Science Foundation
(ESF, Strasbourg, France) Research Network Programme ‘‘European
GlycoScience Forum’’ (EGSF).
References
1 T. W. Rademacher, R. B. Parekh and R. A. Dwek, Annu. Rev. Biochem., 1988,
57, 785.
2 R. B. Parekh, R. A. Dwek, B. J. Sutton, D. L. Fernandes, A. Leung, D.
Stanworth, T. W. Rademacher, T. Mizuochi, T. Taniguchi, K. Matsuta, F.
Takeuchi, Y. Nagano, T. Miyamoto and A. Kobata, Nature, 1985, 316, 452.
3 P. M. Rudd, T. Elliott, P. Cresswell, I. A. Wilson and R. A. Dwek, Science,
2001, 291, 2370.
4 G. Opdenakker, P. M. Rudd, C. P. Ponting and R. A. Dwek, FASEB J, 1993, 7,
1330.
5 P. M. Rudd and R. A. Dwek, Crit. Rev. Biochem. Mol. Biol., 1997, 32, 1.
6 P. Van den Steen, P. M. Rudd, R. A. Dwek and G. Opdenakker, Crit. Rev.
Biochem. Mol. Biol., 1998, 33, 151.
7 D. J. Harvey, Mass. Spectrom. Rev., 2012, 31, 183.
8 J. C. Paulson, J. Weinstein, L. Dorland, H. van Halbeek and J. F. Vliegenthart,
J. Biol. Chem., 1982, 257, 12734.
10 | Carbohydr. Chem., 2012, 38, 1–12

9 H. Lis and N. Sharon, Annu. Rev. Biochem., 1986, 55, 35.
10 J. Jaeken and G. Matthijs, Annu. Rev. Genomics. Hum. Genet., 2007, 8, 261.
11 M. Arakawa, S. Ogata, T. Muramatsu and A. Kobata, J. Biochem., 1974, 75, 707.
12 P. M. Rudd and R. A. Dwek, Curr. Opin. Biotechnol., 1997, 8, 488.
13 Y. Kaneko, F. Nimmerjahn and J. V. Ravetch, Science, 2006, 313, 670.
14 P. M. Rudd, G. R. Guile, B. Küster, D. J. Harvey, G. Opdenakker and R. A.
Dwek, Nature, 1997, 388, 205.
15 K. Mariño, J. Bones, J. J. Kattla and P. M. Rudd, Nat. Chem. Biol., 2010, 6, 713.
16 C. Choudhary and M. Mann, Nat. Rev. Mol. Cell. Biol., 2010, 11, 427.
17 A. Dell and H. R. Morris, Science, 2001, 291, 2351.
18 S. R. Hamilton, R. C. Davidson, N. Sethuraman, J. H. Nett, Y. Jiang, S. Rios, P.
Bobrowicz, T. A. Stadheim, H. Li, B. K. Choi, D. Hopkins, H. Wischnewski, J.
Roser, T. Mitchell, R. R. Strawbridge, J. Hoopes, S. Wildt and T. U. Gerngross,
Science, 2006, 313, 1441.
19 Q. P. Liu, G. Sulzenbacher, H. Yuan, E. P. Bennett, G. Pietz, K. Saunders, J.
Spence, E. Nudelman, S. B. Levery, T. White, J. M. Neveu, W. S. Lane, Y. Bourne,
M. L. Olsson, B. Henrissat and H. Clausen, Nat. Biotechnol., 2007, 25, 454.
20 M. S. Sandrin and I. F. McKenzie, Immunol. Rev., 1994, 141, 169.
21 N. Sharon, Biochem. Soc. Trans., 2008, 36, 1457.
22 W. J. Peumans and E. J. Van Damme, Crit. Rev. Biochem. Mol. Biol., 1998, 33, 209.
23 J. Zhao, T. H. Patwa, D. M. Lubman and D. M. Simeone, Curr. Opin. Mol.
Ther., 2008, 10, 602.
24 R. Saldova, M. R. Wormald, R. A. Dwek and P. M. Rudd, Dis. Markers., 2008,
25, 219.
25 C. D. Rillahan and J. C. Paulson, Annu. Rev. Biochem., 2011, 80, 797.
26 T. A. Springer, Cell, 1994, 76, 301.
27 K. J. Woollard and J. Chin-Dusting, Curr. Pharm. Des., 2010, 16, 4113.
28 B. Lepenies, J. Yin and P. H. Seeberger, Curr. Opin. Chem. Biol., 2010, 14, 404.
29 J. Balzarini, Nat. Rev. Microbiol., 2007, 5, 583.
30 A. Schneider, C. Thiel, J. Rindermann, C. Derossi, D. Popovici, G. F.
Hoffmann, H. J. Gröne and C. Körner, Nat. Med., 2011, 18, 71.
31 T. D. Butters, R. A. Dwek and F. M. Platt, Chem. Rev., 2000, 100, 4683.
32 E. De Clercq, Nat. Rev. Drug. Discov., 2006, 5, 1015.
33 E. C. Borden, G. C. Sen, G. Uze, R. H. Silverman, R. M. Ransohoff, G. R.
Foster and G. R. Stark, Nat. Rev. Drug. Discov., 2007, 6, 975.
34 S. Roy-Ghanta and J. S. Orange, Clin. Rev. Allergy Immunol., 2010, 38, 39.
35 R. Derynck, E. Remaut, E. Saman, P. Stanssens, E. De Clercq, J. Content and
W. Fiers, Nature, 1980, 287, 193.
36 P. De Somer, E. De Clercq, A. Billiau, E. Schonne and M. Claesen, J. Virol.,
1968, 2, 878.
37 P. De Somer, E. De Clercq, A. Billiau, E. Schonne and M. Claesen, J. Virol.,
1968, 2, 886.
38 P. Claes, A. Billiau, E. De Clercq, J. Desmyter, E. Schonne, H. Vanderhaeghe
and P. De Somer, J. Virol., 1970, 5, 313.
39 A. Billiau, J. Desmyter and P. De Somer, J. Virol., 1970, 5, 321.
40 J. Leunen, J. Desmyter and P. De Somer, Appl. Microbiol., 1971, 21, 203.
41 A. Billiau, J. J. Muyembe and P. De Somer, Appl. Microbiol., 1971, 21, 580.
42 A. Billiau, J. J. Muyembe and P. De Somer, Nat. New. Biol., 1971, 232, 183.
43 E. De Clercq and P. De Somer, Eur. J. Cancer., 1972, 8, 535.
44 E. De Clercq, W. E. Stewart and P. De Somer, Infect. Immun., 1973, 7, 167.
45 M. W. Harmon, B. Janis and H. B. Levy, Antimicrob. Agents. Chemother., 1974,
6, 507.
46 O. S. Weislow and E. F. Wheelock, Infect. Immun., 1975, 11, 129.
Carbohydr. Chem., 2012, 38, 1–12 | 11

47 E. De Clercq and M. Luczak, J. Infect. Dis., 1976, 133 (Suppl:A), 226.
48 S. Kluge, Antiviral. Res., 1985, 5, 291.
49 H. B. Levy, J. Duenwald and C. E. Buckler, Infect. Immun., 1973, 7, 457.
50 S. Li, S. Starckx, E. Martens, C. Dillen, N. Lamerant-Fayel, N. Berghmans, M.
Gouwy, M. van Pel, H. Heremans, C. Kieda, W. E. Fibbe, A. Billiau, J. Van
Damme and G. Opdenakker, J. Leukoc. Biol., 2010, 88, 1017.
51 S. Li, E. Martens, C. Dillen, P. E. Van den Steen and G. Opdenakker, Biochem.
Pharmacol., 2008, 76, 831.
52 J. Van Damme, A. Wuyts, G. Froyen, E. Van Coillie, S. Struyf, A. Billiau, P.
Proost, J. M. Wang and G. Opdenakker, J. Leukoc. Biol., 1997, 62, 563.
53 A. M. De Palma, H. J. Thibaut, S. Li, I. Van Aelst, C. Dillen, M. Swinnen, E.
Verbeken, J. Neyts and G. Opdenakker, J. Pathol., 2009, 217, 633.
54 G. Opdenakker and J. Van Damme, Cytokine Growth Factor Rev., 2011, 22, 359.
55 D. Hanahan and R. A. Weinberg, Cell, 2011, 144, 646.
56 H. Piccard, R. J. Muschel and G. Opdenakker, Crit. Rev. Oncol. Hematol.,
2012, 82, 296.
57 H. Piccard, N. Berghmans, E. Korpos, C. Dillen, I. V. Aelst, S. Li, E. Martens,
S. Liekens, S. Noppen, J. V. Damme, L. Sorokin, G. Opdenakker, Int. J.
Cancer, 2012, 131, E425.
12 | Carbohydr. Chem., 2012, 38, 1–12

Lipopolysaccharide structure and biological
activity from the cystic fibrosis pathogens
Burkholderia cepacia complex
Anthony De Soyza,*a Flaviana Di Lorenzo,b Alba Silipo,b
Rosa Lanzettab and Antonio Molinarob
DOI: 10.1039/9781849734769-00013
1 Introduction
This chapter will focus on lipopolysaccharide in cystic fibrosis (CF) related
pathogens.
Cystic Fibrosis is the most common lethal single gene disorder in Europe.
The condition arises due to an autosomal recessive gene defect with an
incidence of approximately 1/2000 live births in the Western hemisphere.
CF has multiple organ system manifestations including liver, gastro-
intestinal, reproductive and pulmonary involvement. Although a multi-
system disease the predominant morbidity and premature mortality relates
to recurrent lung infections that lead to respiratory failure. The basic genetic
defect in Cystic fibrosis (CF) reflects mutations in a gene encoding for an
epithelial chloride channel, named Cystic Fibrosis Transmembrane Reg-
ulator (CFTR). There are now over 1000 described mutations. These cause
abnormal electrolyte flow across the mucosal surfaces resulting in abnor-
mally viscid airway secretions. This in turn prevents normal mucociliary
function and predisposes to chronic bacterial lung infections. Persistent
bacterial infection is a hallmark of the condition with episodic worsening
‘‘exacerbations’’ leading to multiple courses of antibiotics, hospitalization
and precipitous terminal decline. By adulthood the majority of patients with
CF will have chronic low grade infection often termed ‘‘colonisation’’
despite the significant morbidity associated with this chronic infection state.
Importantly colonisation should be seen as a misnomer as there are serious
consequences of these infections with evidence of ongoing inflammation.
We will use persistent infection hereafter this occurs in most patients with
CF by adulthood. Whilst the genetic defect for CF has been understood for
over 20 years as yet gene therapy has not emerged as a clinically useful
therapy.1 Gene therapy is felt most likely to be successful in early stage
disease in the absence of significant pulmonary infection. Therefore there
will still be significant cohorts of patients with CF with established lung
infection requiring improvements beyond our current therapeutic regimens.
Mortality is improving; CF was previously fatal in childhood. Survival is
better but is still limited. The US Cystic Fibrosis foundation notes the mean
age at death for those with cystic fibrosis remains less than 45 years. Even in
the modern era over 80% die of respiratory complications.2
a
Transplantation and Immunobiology group, Institute of Cellular Medicine, Newcastle
University, UK, The Freeman Hospital, High Heaton, Newcastle-upon-Tyne, NE7 7DN,
England, UK. E-mail: anthony.de-soyza@ncl.ac.uk
b
Dipartimento di Chimica Organica e Biochimica, Università di Napoli, Federico II, Napoli,
Italy.
Carbohydr. Chem., 2012, 38, 13–39 | 13

1.1 Cystic Fibrosis pathogens
Characteristic pathogens include Haemophilus influenza, Staphylococcus
aureus3 and in adults the predominant pathogen is Pseudomonas aerugi-
nosa.4 Whilst other pseudomonads can infect patients with CF herein
Pseudomonas aeruginosa will be referred to as Pseudomonas for simplicity.
Important but less common bacterial pathogens include the Burkholderia
cepacia complex (BCC),5 non tuberculous Mycobacteria (NTM), Acineto-
bacter spp,6 Stenotrophomonas spp,7 Achrombacter spp8–10 and Pandorea.
Though some of these can occasionally infect normal human lungs these
pathogens are predominantly opportunistic pathogens. Despite this pre-
dominance of opportunistic pathogens, a surprising variety of bacterial
toxins, proteases, lipases and other virulence factors are produced by the
pathogens of interest.11,12 A common clinical pattern is a sequential
replacement of pathogens over time from Haemophilus and S. aureus with
Pseudomonas usually becoming the predominant pathogen by adulthood.13
When compared to other CF pathogens persistent Pseudomonas chronic
infection has been associated with more rapid decline in health status.4,14–16
An individual will usually harbour a unique clonal strain over a period of
many years. There are however reported outbreaks of epidemic strains
within CF clinics particularly noted for Pseudomonas,17–19 Burkholderia20
and Staphylococcus aureus.21 Strict infection control mechanisms appear to
greatly reduce these outbreaks.
Selective pressures on bacteria in persistent chronic CF infection are
plausible with exposure to repeated antibiotics, continual exposure to host
defence molecules and changes in the local environment caused by wor-
sening lung disease all likely forces.22,23 There is a widely held but poorly
defined concept that ‘‘early’’ stage strains isolated soon after acquisition
may have different characteristics to those seen in persistent ‘‘late’’ stage
strains.24,25 The definition of early stage and late stage is complex and
somewhat arbitrary. A recent excellent review covers this area well.23 It is
still somewhat unclear if the adaptation takes years or could be occurs
within months although some data from infants with CF suggest that
changes occur very rapidly.23 It is biologically plausible that the char-
acteristics required for acute infection such as adhesion and invasion are
likely to be different to those required to maintain persistent infection.
Certain themes emerge in Pseudomonas aeruginosa strains from patients
with advanced CF lung disease; antimicrobial resistance rates including
pan-resistant strains are highly prevalent. Furthermore development of
alginate and the ‘‘mucoid phenotype’’ and a hypermutator phenotype are
also common during chronicity of infection.26 Biofilm forming capacity is
common in strains isolated cystic fibrosis particularly those who have per-
sistent infection for years.27,28 There is usually an increased rate of anti-
microbial resistance in biofilms28,29 thereby further compounding the
antimicrobial resistance profiles seen in planktonic state.
1.2 Cystic Fibrosis as an inflammatory disorder

CF is characterised by an inflammatory airway syndrome with mucus
retention, neutrophilic airway wall inflammation and permanent dilatation
of the airways (bronchiectasis). Retention of secretions further promotes
14 | Carbohydr. Chem., 2012, 38, 13–39

bacteria persistence and prevents mucociliary clearance of virulence factors.
These processes collectively promote further airway damage and inflam-
mation leading to the vicious cycle paradigm. Recently CFTR has been
shown to be a negative regulator of a central pro-inflammatory controller
Nuclear Factor kappa B (NF-kB).30 Hence mutations in CFTR may be in
themselves pro-inflammatory. Cytokines of relevance in CF include inter-
leukin 8 (IL-8), IL-1, IL-6, IL-17 and TNF-alpha.31–33 These cytokines have
a pro-inflammatory effect through vasodilatation, direct chemokinesis,
neutrophil priming and angiogenesis. Recently more novel pro inflamma-
tory moieties that have been associated with inflammation in CF include
endogenous damage associated molecular patterns (DAMPS) such as high
mobility group box 1 (HMGB1)34 and lipid metabolites such as ceramide.35
Despite elevated numbers of airway innate immune cells and evidence of
microbe specific adaptive immunity there is clearly impaired immunity.36
There are data noting neutrophil dysfunction in CF with evidence of neu-
trophil derived peptides (HNP) inhibiting neutrophil activity.37 This may
appear counterintuitive that neutrophils release a peptide that inhibits their
own function- the observed effect may reflect the inhibitory activity at high
local concentrations. There are also studies suggesting bacteria may cause
human receptor dysfunction such as inhibiting CXC chemokine receptor
biology.38 Hence CF is characterised by high levels of cytokines and
chemokines that promote vascular dilatation, influx of inflammatory cells
and release of noxious defence molecules that may have bactericidal activity
but at high concentrations lead to the perpetuation of host damage.
1.3 Cystic Fibrosis as a complex microbial community disease

The conventional paradigm for acute infections arising from a single
pathogenic organism still stands e.g. acute pneumonia caused by Strepto-
coccus pneumonia. In contrast chronic infection in cystic fibrosis may need a
newer paradigm. The poor reproducibility of CF lung disease using agar
beads and Pseudomonas in animal models in part supports this; a non-acute
infection can be established but rarely persists with CF type lung pathology.
Increasing evidence now supports a concept of polymicrobial infection as
evidenced by non-culture based diagnostics and markers of biofilm infec-
tion. Non culture based techniques have recently shown a high diversity of
organisms in cystic fibrosis.39–41 Many pathogens or potential pathogens
have been identified through such techniques though the direct clinical
relevance is unclear. Certainly many of those identified in this manner are
inherently un-cultureable either through their obligate intracellular
requirements or need for anaerobic conditions42 which cannot be replicated
in a routine diagnostic microbiology lab. Biofilms are complex communities
of bacteria with physical and functional aspects. Biofilms can be seen in the
CF lung that contributes to pathogenesis by increasing resistance to pha-
gocytosis, increased rates of antibiotic resistance28,43,44 and is likely to
contribute to persistence of organisms. Good in vitro data supports many
CF pathogenic organisms can form biofilms and that they can do so in
monoculture and in complex cultures.29 The importance of multispecies
biofilms is that supposedly low pathogenicity bacteria such as oral ‘‘com-
mensal’’ can induce virulence factor upregulation in Pseudomonas
Carbohydr. Chem., 2012, 38, 13–39 | 15

aeruginosa. The effect of complex biofilm growth on cell surface virulence
determinants is to date still an emerging field.
2 Basic Lipopolysaccharide structure

Lipopolysaccharides (LPS) are the major components of the outer mem-
brane of Gram-negative bacteria and exert, in isolated form, a variety of
biological activities in animals and plants.45,46 From the structural point of
view, LPSs are amphiphilic macromolecules that show the same general
chemical architecture in all of the Gram-negative bacteria, that is, it does
not depend on the bacterial activity, (pathogenic, symbiotic, commensal),
nor on its ecological niches: (human, animal, soil, plant, water) or on its
growth conditions. They are indispensable for growth and survival of
Gram-negative bacteria and comprise about 10–15% of the total molecules
in bacterial outer membranes covering about 75% of bacterial surface area,
thus they are exposed to the bacteria’s external environment. LPS represent
a defensive barrier which helps bacteria to resist to antimicrobial com-
pounds and environmental stresses and are involved in many aspects of
host–bacterium interactions as recognition, adhesion, colonization. Once
LPS molecules are released from dead bacteria they play a key role in the
pathogenesis of Gram-negative infections potentially causing fever or cir-
culatory shock. LPS has additional key roles in virulence, tolerance for
commensal bacteria and symbiosis.45 From both genetic and chemical
viewpoints, LPSs have a common structural architecture consisting in three
domains, (Fig. 1): a polysaccharide (known as O-side chain, O-specific
polysaccharide or O-antigen) which is covalently linked to an oligo-
saccharide part (core), in turn, linked to a glycolipid portion (Lipid A).
The lipid A component of LPS is part of the outer leaflet of the external
membrane whereas the sugar moiety is oriented outwards.47 Bacterial
colonies can be catalogued by gross colony morphology as rough or smooth
which is related to the occurrence of the O-side chain, being absent in the
former and present in the latter. The terminology presently used to desig-
nate the different LPS types, namely with or without the O-polysaccharide
Fig. 1 Scheme of a general chemical structure of bacterial lipopolysaccharides.
16 | Carbohydr. Chem., 2012, 38, 13–39

antigenic portion, is S-LPS or R-LPS (or lipooligosaccharide, LOS),
respectively.
In all cases, the LPS fraction of an individual bacterial strain is char-
acterised by an intrinsic size and structural heterogeneity. Hence it is more
appropriate to view LPS as a blend of various LPS molecules (and similarly
so for LOS).
2.1 Structural investigation of the glycolipid (LPS or LOS)

LPSs (or LOSs) may be extracted from intact cells by a variety of proce-
dures, but currently those commonly used exploit the phenol/chloroform/
light ether and/or phenol-water extractions. Depending on the polarity of
glycolipid it can be isolated from aqueous, phenol or the organic phase.
Enzymatic treatments that remove contaminating proteins and nucleic
acids and chloroform washings that remove lipid and phospholipids are
frequently used. The amphiphilic nature of these molecules is responsible
for their low solubility in both aqueous and organic solvents and renders the
purification by any chromatographic technique problematic. In contrast
electrophoretic analysis using a denaturing agent is highly informative and
allows chemotyping of the extracted material: a ladder like profile is diag-
nostic of LPS molecules reflecting the O-polysaccharide. In contrast the
presence of material at bottom of the gel profile, without the ladder profile,
denotes the smaller molecule LOS lacking O-polysaccharide. This infor-
mation is valuable and affects the subsequent procedures for the structural
investigation.
However, in both cases, the study of the lipid A moiety is carried out after
its cleavage from the rest of the molecule by mild acid hydrolysis of the acid-
labile glycosydic linkage of Kdo. During this treatment, the lipid A part
precipitate whereas the saccharide moiety remains soluble in the supernatant.
The primary structure determination of very complex molecules such as
LPSs (LOSs) currently exploits state-of-art 1D and 2D NMR experiments
and soft-ionization mass spectrometry techniques, together with composi-
tional and linkage chemical analyses, usually carried out on intact products
(when possible) and ad hoc selectively degraded derivatives.
2.2 Lipid A structure elucidation

The lipid fraction obtained by precipitation, after mild acid treatment of
LPS (or LOS) consists of a mixture of intrinsically heterogeneous lipid A
molecules. The structural approach makes an extensive use of mass spec-
trometry analysis by MALDI-TOF (Matrix-Assisted Laser Desorption/
Ionization – Time Of Flight mass spectrometry) and ESI MS (Electrospray
Ionization mass spectrometry) techniques both on the native and selectively
degraded lipid A fractions. The MALDI-TOF and ESI-MS data allow
gaining insights into the number of lipid A species present in the fraction,
the presence of polar heads and the distribution of acyl residues on each
GlcN units of the disaccharide backbone.48 This latter issue is of particular
importance given that fatty acid composition and distribution strongly
affects lipid A/LPS endotoxic activity, i.e. agonist/antagonist activity.
Through the chemical and spectroscopical study of several lipid As and
LPSs in our laboratory, novel general approaches to gain structural insights
Carbohydr. Chem., 2012, 38, 13–39 | 17

on such molecules have been developed. In particular, combining MALDI-
MS or ESI-MS and selective chemical degradation by ammonium hydro-
xide hydrolysis, we have developed a general and practical methodology to
obtain the secondary fatty acid distribution, which is one of most
demanding issues in the structural determination of lipid A.49,50,51 These
approaches must be supported by the classical chemical analyses of the lipid
A, which imply the removal of ester-linked fatty acid moiety by mild
alkaline treatment, or complete removal by strong alkaline treatment and
GLC-MS analyses of fatty acids as methyl ester derivatives.52,53
The NMR study of LPS is less frequently used but is equally impor-
tant54,55; the amphiphilic nature of lipid A entails a very poor solubility of
the sample in whichever solvent system, however, when possible it is
invaluable in obtaining diagnostic information. NMR is highly useful for
studies on the saccharide backbone, such as the anomeric configuration of
the carbohydrate residues and their sequence, the occurrence and position
of the polar heads.
2.3 O-chain polysaccharide structure

The O-chain of LPSs, recovered in supernatants after mild acid treatment,
usually consists of a regular long chain polysaccharide with the core oli-
gosaccharide still at the reducing end of the chain, but, since this oligo-
saccharide moiety represents a small percentage of the whole polymer, it is
generally barely detectable from the chemical and spectroscopical point of
view. The O-chain structure determination aims to define the composition
in monosaccharides, the relative, the anomeric and the absolute config-
urations of each monosaccharide, their ring sizes, the linkage of each residue
and their sequence within the repeating unit. In some cases, the non-stoi-
chiometric presence of some substituents can determine heterogeneity and
hide the structural regularity.56
The monosaccharide composition is achieved by the classical analytic
approaches consisting in acid hydrolysis and/or methanolysis followed by
further chemical modification (such acetylation, for example) to perform
GLC-MS analyses. The reaction sequence: methylation, acid hydrolysis,
reduction and acetylation are carried out to know the glycoside substitution
sites and the ring sizes of each monosaccharide. For the absolute config-
urations, GLC analysis of convenient glycoside derivatives are usually
exploited. The anomeric configurations and the sequence of the mono-
saccharide residues are determined by 1D and 2D 1H and 13C-NMR
experiments (COSY, TOCSY, NOESY, ROESY, HSQC, HMBC). In
detail, TOCSY and DQF-COSY spectra allow the correct identification and
assignment of all ring protons, and on the basis of these data, the assign-
ment for all 13C resonances usually follows from the analysis of an 1H,13C-
HSQC spectrum; the intra-residue pattern of dipolar correlations in the 2D
NOESY or ROESY spectra gives the proof for specific configurations.
Information concerning the sequence of the monosaccharides within the
oligosaccharide (or the repeating unit of the polysaccharide) are then
deduced from the observation of the inter-residual dipolar correlations in
2D NOESY and ROESY spectra and from the existence of scalar long
18 | Carbohydr. Chem., 2012, 38, 13–39

range correlations, in the 1H,13C-HMBC spectrum, among the 1H and 13
C
nuclei involved in glycosidic linkages.
2.4 Core oligosaccharide structure

The target of the structural elucidation of a LOS core oligosaccharide is
generally much more complex than that of the O-chain because of the
higher monosaccharide number, commonly up to 15, with respect to the
components of one O-chain repeating unit, and the presence of labile groups
and other substituent that are often present in non stoichiometric amounts.
In addition, the acid hydrolysis used to remove the lipid A from LOS occurs
by hydrolysis of Kdo that in its reducing form usually appears in different
anomeric and cyclic forms, thus determining an ‘‘artificial’’ increase of the
core heterogeneity. Therefore, LOS sample is generally treated under
alkaline conditions that do not cleave the Kdo linkage thereby allowing the
analysis of the complete sugar backbone including the two GlcN of lipid A.
Notably the LOS sample should be first treated under mild alkaline con-
ditions in order to selectively remove the lipid A ester-bound fatty acids.
This more soluble O-deacylated LOS is N-deacylated under more strong
alkaline conditions affording completely deacylated backbone oligo-
saccharide, which represents the intact sugar skeleton of LOS suitable for
sequence determination by NMR analysis.57 The direct strong de-O and
de-N-acylation is usually not performed in order to avoid the formation of
alkali resistant unsaturated N-linked fatty acids. In order to assign phos-
phate groups and their possible further attached substituents 1D and 2D
31
P-NMR experiments are also carried out.
Since alkaline treatment frequently determines the loss of base-labile
functions, the oligosaccharide moiety released under acid conditions is also
studied as a complemetary and informative approach. More recently, the full
NMR study of the O-deacylated sample has been carried out in a water
solution containing a denaturing agent. Every step of the selective degradation
is usually accompanied by MALDI MS spectrometry analysis that is a very
valuable and complementary approach in the study of such molecules. Our
collaborators have pioneered the development of new MALDI MS approa-
ches through development of new methods of preparation and experimental
study. It is worth emphasising that when such experiments are carried out on
the intact sample they are often of very high value as they yield a ‘‘picture’’ of
the intact endotoxin without any loss of structural information.58
2.5 Relevance of structure

The importance of LPS structure is made eloquently clear in the clinical
management of CF. A commonly used therapy includes polymixin E or
colistin; this cationic antimicrobial peptide has significant microbicidal
activity against many CF pathogens and attacks the bacterial cell mem-
brane through a physicochemical interaction with LPS. This interaction is
determined by the LPS composition and surface charge. Hence the majority
of CF isolates of Pseudomonas have negligible rates of clinically relevant
colistin resistance whilst Burkholderia species are to date almost always
resistant. Further pathogens in CF such as the enterobacteriae Serratia and
Proteus are also often resistant to colomycin which again reflects structural
Carbohydr. Chem., 2012, 38, 13–39 | 19

composition of the lipid A as discussed below. These resistance patterns
dictated by lipid A composition are reproducible enough to have some
diagnostic value: If clinical diagnostics laboratory detects a Pseudomonas
isolate with high level resistance to colistin- a repeat microbial species-
identification step is often undertaken as there such low rates of clinically
significant colistin resistance in Pseudomonas. Often on closer inspection the
organisms has proved to have been misidentified and is not a Pseudomonas
species. Eloquent experiments in knockout models have demonstrated that
the structure of LPS particularly the lipid A component has a marked effect
on the pro-inflammatory activity in vitro. For example the loss of a lipid
A single acyl side chain in Neisseria, a non CF pathogen, can reduce
pro- inflammatory activity in vitro models by up to 50%. Similar data for
E. coli suggests the importance of lipid A acylation patterns across a diverse
range of pathogens.59 Hence LPS structure may dictate two important
clinical end points: resistance to antimicrobial therapy and pro- inflam-
matory activity.
Recent reviews have comprehensively studied Pseudomonas LPS and the
reader is directed to these.60,61 We will briefly highlight important points.
3 The detailed chemical structure of the P. aeruginosa LPS

Pseudomonas aeruginosa has a typical Gram-negative bacterial LPS, with a
basic lipid A structure containing an N- and O-acylated diglucosamine
bisphosphate backbone [4-P-a-D-GlcpNII-(1-6)-a-D-GlcpNI-(1-P] with
chemical variation in the number of primary acyl groups and the types of
fatty acids substituting the primary and secondary acyl groups (Fig. 2).
Most of the laboratory derived Pseudomonas strains studied so far syn-
thesize a penta-acylated (75% of the molecules) LPS, with some proportion
made as a hexa-acylated LPS (25% of the molecules). The difference
between the 2 isoforms is the lack of an O-linked 3-hydroxy decanoic acid
(10:0 (3-OH)) group at position 3 of the first glucosamine in the penta-
acylated isoform (Fig. 2). Certain Growth conditions e.g. low magnesium
levels, can result in altered P. aeruginosa lipid A acylation patterns. Pseu-
domonas isolated from chronically infected CF patients are often unable to
synthesise O-antigen side chains; these strains have a predominant hexa-
acylated LPS. In certain cases a hepta-acylated lipid A has been isolated,
containing an additional palmitoyl (16:0) group (Fig. 2) linked to the pri-
mary 3-hydroxy decanoic acid group at position 3 0 of glucosamine 2.62 The
hexa- and hepta-acylated lipid A moieties also contain cationic 4-amino-4-
deoxy-L-arabinose sugars. Bound to the lipid A is a relatively conserved
inner-core structure which contains two D-manno- oct-2-ulosonic acid resi-
dues (KdoI and KdoII) and two L-glycero-D-manno-heptose residues (HepI
and HepII).63 Bound to the second heptose residue, HepII, is a 7-O-car-
bamyl group,64 which is found in the LPS of other pseudomonads.65 The 2
heptose residues are often phosphorylated at positions 2 and 4 of HepI and
position 6 of HepII.66,67 Phosphate substituents can be mono-, di- or even
tri-phosphates, with most analyzed P. aeruginosa LPS having some tri-
phosphate, which, to date, has only been detected in the LPS of this bac-
terial species. Ethanolamine mono- or diphosphate is present in some
20 | Carbohydr. Chem., 2012, 38, 13–39

Core Core
O O
O O O O
O P O O O O P O HO
HO
OH O OH O HO
HO O NH NH O
O NH NH O O
O O O O O PO
HO O PO O
O O O O OH
OH O O
O O HO OH
OH
0
10
10
12 12
12 12
12
A 12 B
16 12
Penta 12 Esa1
Core
Core O
O O
O O P O O O
O P O O O HO
HO
OH O O
OH O O O NH O NH O
O NH O NH O O O
O O O HO O PO
HO HO O PO O O
O O O OH
OH O
O HO O OH
O OH
10
10 10 10
12 12
12 12
12
C 12 D
16 12
12 Hepta
Esa2
Fig. 2 Chemical structure of P. aeruginosa lipid A from clinical strains isolated at the onset of
chronic colonization (A is the structure after 6 months of colonization) and after years of
chronic colonization (B, C and D are after 7.5 years of colonization by a mucoid strain and a
non mucoid strain respectively).77
Pseudomonas LPS inner cores in non stoichiometric amounts. Notably a

mutation in a waaP gene that phosphorylates HepI groups at position 4 is
lethal strongly suggesting that phosphorylation is critical for bacterial
survival.68 The outer core of the P. aeruginosa LPS is usually synthesized as
2 different isoforms or glycoforms by an individual strain.69 Both outer-core
glycoforms contain an N-alanylated galactosamine residue, 3 D-glucose
residues, and one L-rhamnose residue the position of which differs in the 2
glycoforms. There is also some structural variation among the different P.
aeruginosa strains in glycoform I, with a less common variant having four
glucose residues in the outer core.69 There is extensive O-acetylation of the
hydroxyl groups in the outer-core sugars, but the acetates are not present in
high amounts at any one position, making a clear structural determination
of these substituents difficult.
As O-acetyl groups are relatively labile under mild acid or basic condi-
tions they may be lost during LPS purification. If the terminal rhamnose is
not substituted with another monomonosaccharide it is often acetylated.
With the exception of O-antigens from O14 and O15 serogroups most P.
aeruginosa LPS-smooth strains studied to date have glycoform II fully
substituted with an O-antigen. LPS-rough strains, usually isolated from CF
patients, have multiple mutations within the biosynthetic genes for the
O-antigen70 and do not have any O-antigen on glycoform II.71 The
O-antigen portion of the P. aeruginosa LPS is responsible for conferring
serogroup specificity, which is defined by antibodies specific to the different
Carbohydr. Chem., 2012, 38, 13–39 | 21

variants of this antigen. A variety of serotyping systems have been proposed
in the past; most of these can be reconciled when cross-referenced with prior
data and strains.70 In general, most investigators now refer to serogroups
using the International Antigenic Typing System (IATS) schema. Chemi-
cally the O-antigens can be categorized into at least 11 structural variants,66
although within these variants there are minor differences among related
structures in properties such as sidegroup substituents, linkages between
sugars or conformations of different monosaccharides.66 P. aerguinosa
serogroups O14 and O15 strains both lack detectable O-antigens or LPS
containing even a single O-antigen repeat linked to the glycoform 2 outer
core (referred to as the SR-type LPS). This may point to the fact they
probablyy LPS-rough bearing strains in spite of the availability of ser-
ogroup-specific antisera. It is pluasible that the antisera to the serogroup
O14 and O15 antigens are cross reacting with LPS core epitopes or other
components in these serogroup strains. Typical sugars within the P. aeru-
ginosa LPS O side chains include N-acyl derivatives of different amino
sugars along with rhamnose.71 The structures of some of these mono-
saccharides were first described in P. aeruginosa LPS. Rhamnose is also
commonly found in some of the O-antigens. The monosaccharides are
arranged in repeat units containing 3 to 4 individual monosaccharides,
except for serogroup O7, which is a disaccharide repeat unit. The linkage
of the monosaccharide in the first repeat unit to the rhamnose residue
in glycoform II of the LPS outer core is usually a 2-N-acetyl derivative of
a 6-deoxy-D-hexosamine, usually D-quinivosamine or D-fucosamine,
although for serogroup O3 it is the 4-amino, 4-deoxy derivative of
D-quinivosamine.
3.1 Structural adaptation of LPS to the host CF airway

Selected Gram-negative bacteria have evolved mechanisms to modify the
structure of lipid A in different environments, including in host tissues.
There is evidence that bacterial resistance mechanisms, such as lipid A
modifications may have evolved to protect against host cationic anti-
microbial peptides (CAMP).72 The mucosal epithelial surface produces a
number of antimicrobial peptides. Hence alteration of the surface charge of
bacteria provides a defence mechanism for the bacteria against these host
derived CAMPs.73 The systems for evoking these changes are only partly
understood and in Pseudomonas appear to involve a two component mag-
nesium dependent sensor system PhoP/PhoQ.74–76 This mechanism does not
appear present in all CF pathogens with no gene homologues easily iden-
tified in BCC, for example (De Soyza, unpublished observations). Selected
CF pathogens are associated with specific chemical alterations in their
outer membrane when the strain has become chronically established.
P. aeruginosa clinical bacterial isolates from CF have a unique lipid A
structure.62 These lipid A molecules are structurally different from those
that are found in environmental isolates and isolates from other human
diseases including acute urinary, blood or eye infections.66 Lipid A from P.
aeruginosa CF isolates, in contrast to isolates from other human infection
sites, markedly stimulate human Toll receptor 4 (TLR4) signalling and
22 | Carbohydr. Chem., 2012, 38, 13–39

might contribute to the inflammatory response that is seen in patients with
early stage CF.66 Early adaptation of P. aeruginosa to the CF airway by
synthesis of a modified lipid A, that includes the addition of aminoar-
abinose and palmitate, leads to resistance to host innate immune defences
(such as cationic antimicrobial peptides) and increased pro-inflammatory
signalling. These modifications appear stable and were not lost after con-
tinual passage in rich medium in vitro. These results suggest that lipid A
modifications are constitutively and stably expressed in P. aeruginosa iso-
lates from CF patients after growth in vitro.66 In general, production of fully
hexa-acylated lipid A as seen in CF is associated with a more strong TLR4
mediated inflammatory response while lipid A with lower levels of acylation
triggers reduced cellular responses. in contrast to early stage CF infections
there is a switch to penta- and hexa-acylated lipid A in later stages of CF.
Therefore the LPS from a P. aeruginosa strain isolated from a patient with
late stage CF lung disease may induce less inflammation than LPS from an
early isolate from the same individual. Such LPS modifications in CF may
be a form of host immune evasion.77 Importantly there is a species specific
effect whereby human TLR4 signalling is altered by these chronic strains
‘‘CF lipid A’’ changes whilst mouse TLR4 did not respond differentially to
early stage and late stage CF lipid A.78 This has led to the concept that the
bacteria adapts specifically to the host TLR.77 Recent data has suggested
CF strains not only differ in their potency to induce inflammatory responses
as compared to environmental strains but they also have differing cytokine
induction profiles. Divergent responses were seen when comparing ‘‘early
stage’’ CF infection strains to ‘‘late stage’’ strains and an non clinical strain
PAO1: The LPS from strain PAO1 was a more effective inducer of IL-8 than
any of the early or late stage CF strains but elicited less TNF-alpha from
epithelial cells.77
Previously reviews have covered important aspects of the important
virulence determinants BCC lipid A structure, O polysaccharide and link
these to biological activity.46,79 We will highlight relevant newer data and
where contextually important re-address selected elements of data pre-
viously covered.
4 Burkholderia cepacia complex and Cystic Fibrosis

‘‘Burkholderia cepacia’’ has long been identified as a plant pathogen where it
causes onion rot.80 More recently it has been recognized as an opportunistic
human pathogen for individuals with cystic fibrosis (CF)81 and also those
with chronic granulomatous disease.82 Previously, infections with ‘‘Bur-
kholderia cepacia’’ were attributed to a single species. Indeed, such infec-
tions are caused by a variety of bacterial species, grouped into the
‘‘Burkholderia cepacia complex (BCC)’’.20 Currently this complex comprises
over 10 related species named genomovars which are phenotypically similar
but genotypically different. The use of formal species names is now pre-
ferred although to date there are no phenotypic tests that separate each
genomovar.
The most prevalent clinical isolates from CF patients are B. cenocepacia
strains (genomovar III) and B. multivorans strains (genomovar II)83
Carbohydr. Chem., 2012, 38, 13–39 | 23

and infection with the former is most often associated with patient-to-
patient spread and fatal invasive infection (the so-called ‘‘cepacia
syndrome’’).11
Lung transplantation, whilst intensive and aggressive, is to date the only
management strategy that offers improved quantity and quality of life of CF
patients with advanced lung disease; BCC’s impact on the survival of CF
patients following lung transplantation is viewed as so detrimental that
infection with these organisms is considered a relative or absolute contra-
indication for transplantation.84,85 Post-transplant survival rates are 15% to
63% lower than those of CF patients not infected with BCC,86 with death
usually resulting from BCC sepsis in the early postoperative period.84 Not
all BCC species cause poor outcomes following lung transplantation.86–89
Furthermore, there are great differences in vitro among different isolates
from the same genetic lineage e.g. for the renowned B. cenocepacia epidemic
strain ET-12 clonal strains J2315 and K56-2 strains90 behave differently
when used in vitro and in vivo models. Hence this heterogeneity illustrates
the importance of examining multiple clinical bacterial isolates and not
generalizing the results from studies of a single strain.
Molecular differences among the BCC family has been suggested as the
cause of their variable pathogenicity, and their LPS has been proposed as
potentially important in this. Studies conducted soon after genomovar
designation have assessed LPS chemotypes within the BCC.91 So far in our
labs all strains tested belonging to B. cepacia genomovar I and B. multi-
vorans had smooth LPS. In contrast, isolates belonging to B. cenocepacia,
expressed either rough or smooth LPS chemotype.92,93 Importantly two
clones of the B. cenocepacia ET 12 lineage differed with J2315 and K56-2
having a rough and a smooth chemotype respectively. Subsequent studies in
clinical BCC strains assessing chemotype have demonstrated both rough
and smooth LPS chemotypes are found. Other genomovars studied had
predominantly smooth LPS though B. multivorans and B. vietnamiensis
strains expressing rough LPS were noted.94 Based upon historical studies
using rabbit antisera the LPS core of the strain B. cenocepacia J2315
appears antigenically related to some but not all other B. cenocepacia strains
(reviewed in46,79).
Interestingly very limited cross reactivity, between B. cepacia genomovar
I LPS and no cross reactivity with Pseudomonas LPS was observed when
probed with antisera taken from patients suffering from B. pseudomallei
infection using a microbead-LPS conjugate based assay. The anti-sera from
B. pseudomallei infected patients did however cross react with more phy-
logenetically related bacteria such as B. mallei and B. thailandensis (one
previously called arabinose-negative B. pseudomallei).95 Collectively these
data suggest distinct antigenic clustering within the Burkholderia genus
with similarities between BCC genomovars but not other Burkholderia
species or Pseudomonas. Readers should be aware of the importance of
growth conditions in vitro: Variation in antigenic cross reactivity and LPS
phenotypes has been reported in a comparative study of BCC bacteria
grown in either solid agar or broth. Hence certain BCC strains may dif-
ferentially express LPS dependent on the local environment. This is clini-
cally important as airway mucus may serve as a semi-solid growth media
24 | Carbohydr. Chem., 2012, 38, 13–39

within CF lungs. As most published studies including our own have
extracted LPS from bacteria grown in broth a careful appraisal of these
prior studies is needed.
4.1 General structural features of the lipid A of BCC LPS

According to early compositional data of Cox et al., all the lipid A struc-
tures of the Burkholderia strains elucidated contain a general architectural
in which the following elements are always present: the two 2-amino-2-
deoxy-glucopyranose (GlcN) carbohydrate backbone and an additional
carbohydrate element, the 4-deoxy-4-amino-arabinopyranose (Ara4N); the
phosphate groups; the C16(3-OH) and C14(3-OH) as primary fatty acids and the
C14:0 as secondary fatty acid. These elements are very often present in a non
stoichiometric fashion giving rise to a wide array of lipid A primary
structures that varies on the strain and the environment from which the
bacteria has been isolated. As for fatty acids, the above moieties are the
most abundant component found in the BCC lipid As, but in selected
strains and in particular conditions other minor components have been
found, such as C12:0 (rather than C14:0) or C14:(2-OH) fatty acids.46 As yet the
only isolated structures of BCC lipid A are penta- or tetra-acylated with no
formal identification of hexa-acylated structures. All these data confirmed
previous studies noting a resistance of BCC strains to host defence peptides
different to that of Pseudomonas which suggested divergence in the lipid A
and LPS structure between BCC strains from that of Pseudomonas species.
Indeed, various BCC strains, including three CF isolates and two reference
strains, were more resistant to killing by protegrin-1 (PG-1), a potent
antibiotic peptide able to prevent infection via attack on the bacteria
membrane surface. Furthermore, binding/diffusion and surface plasmon
resonance assays revealed that isolated lipopolysaccharides and lipid As
from the sensitive P. aeruginosa strains bound PG-1 more effectively than
LPS and lipid A of resistant BCC strains.46
Although a body of publications related to the biology and biochemistry
of lipid A/LPS from BCC has appeared in recent years, scanty data are
available on the molecular structure of these important biomolecules.
In this chapter we report the general features of BCC strains lipid A
characterized to date.
4.2 Burkholderia cepacia genomovar I

The primary structure of the lipid A of B. cepacia genomovar I has been
recently elucidated96 by chemical degradations, mass spectrometry and
NMR spectroscopy. These analyses revealed a mixture of species differing in
the acylation and in the phosphorylation patterns. The most frequent spe-
cies was found to be a penta-acylated tetrasaccharide backbone with two
phosphoryl-arabinosamine residues in addition to the archetypal glucosa-
mine disaccharide[Arap4N-L-b-1-P-4-b-D-GlcpN-(1-6)-a-D-GlcpN-1-
P-1-b-L-Arap4N].96 Lipid A fatty acids substitution was also noted, with
two 3-hydroxytetradecanoic acids 14:0 (3-OH) in ester linkage, and two 3-
hydroxyhexadecanoic acids 16:0 (3-OH) in amide-linkage, one of which was
substituted by a secondary a 14:0 residue at its C-3.
Carbohydr. Chem., 2012, 38, 13–39 | 25

4.3 Burkholderia multivorans genomovar II and B. vietnamiensis
genomovar V
Both B. multivorans and B. vietnamiensis possess an equally heterogeneous
lipid A in which tetra-acylated and penta-acylated species (lacking or not a
primary 14:0 (3-OH) in ester linkage to the GlcNI) coexist in nearly equal
amounts. The lipid A in both these genomovars can bear up to two Ara4N
residues with phosphate groups also present in non stoichiometric amount.
4.4 Burkholderia cenocepacia genomovar III

The B. cenocepacia genomovar III ET-12 clone (J2315) lipid A has a rela-
tively marked heterogeneity in fatty acid content with a consistent pro-
portion of tetraacylated species and only one Ara4N residue at anomeric
position of GlcNI.93 This contrasts with the strain B. cenocepacia
LMG12614, another ET-12 clone, which possesses a much more homo-
geneous lipid A in which there is a major penta-acylated species with one or
two Ara4N residues. Interestingly our early data noted many shared lipid A
species between B. cenocepacia LMG12614 (an ET-12 clone) and B. mul-
tivorans LMG14273 though there was a greater proportion of penta–acy-
lated species in the B. cenocepacia strain.93
Thus far, the fully established lipid A structures of BCC LPS consist of a
blend of pentacylated and tetracylated structures (lacking a primary 14:0
(3-OH)) in which two phosphoryl-arabinosamine residues can be present in
non stoichiometric amount.46 These chemical structures have strong simi-
larities with other non BCC Burkholderia spp. lipid A, as for example the
plant pathogen B. caryophylli LPS46,97 in which the lipid A is nearly iden-
tical to the one of B. cepacia LPS (Fig. 2). Collectively these data suggest the
biosynthetic pathways for lipid A may not be unique to a particular gen-
omovar but differential lipid A acylation and glycosylation by Ara4N may
depend on environmental pressures and possibly genomovar specific
responses to such pressures (Fig. 3).
4.5 Importance of lipid A structure of BCC strains in CF

Both lipid A acylation and substitution by aminoarabinose (Ara4N) are
pivotal in the survival of Gram negative bacteria against host response.98
Ara4N substitution in lipid A and LPS inner core contributes to the overall
net charge surface of the external membrane and plays a key role in
pathogenesis. In fact, LPS from the Burkholderia genus are frequently
positively charged or in an isoelectric state. This is closely related to the
abundance of Ara4N residues present in the lipid A-inner core and confers
resistance to antibiotic compounds particularly host cationic antimicrobial
peptides. In physiological conditions Ara4N moieties are positively charged
which reduces the net surface charge on the bacterial external membrane
weakening the ionic attraction for respiratory tract antimicrobial defensins.
The inherent resistance of the BCC micro-organisms to polymyxin B, a
cyclic polycationic antibiotic peptide that has highly affinity for negatively
charged bacterial LPS, is likely to be Ara4N dependent. These residues
prevent the antimicrobial action of polymyxin B, namely increasing per-
meability in bacterial outer membrane. Interestingly, the other frequent
cationic binding groups (e.g., ethanolamine) have not been reported in lipid
26 | Carbohydr. Chem., 2012, 38, 13–39

OH
O
O O O NH2
P O
OH O O HO O OH
HO HO O O
NH2 NH NH P O HO
O O O O
O O
OH
OH O OH OH
O
14
14
16
14 16
Fig. 3 Lipid A from B. cepacia genomovar I. The structure of the different lipid A molecules
from B. cepacia LPS sketched in a single formula. The dotted lines indicate non stoichiometric
substitutions
A from B. cepacia, suggesting a unique role for L-Ara4N.98 Hence, the

transferase enabling the attachment of L-Ara4N residues to lipid A and
whichever earlier implicated enzymatic step may be a potential novel target
for the future developments of drugs against B. cepacia. The gene cluster
encoding aminoarabinose (Ara4N) biosynthetic enzymes is essential for
Burkholderia cenocepacia viability. In a conditional mutation strategy using
rhamnose supplemented media, marked alteration of bacterial cell perme-
ability, osmotic sensitivity, gross colony morphology and ultrastructure was
seen when aminoarabinose biosynthesis was disrupted. These data suggest a
fundamental role for Ara4N in maintaining normal cell membrane structure
and function in BCC. The biological activity at inducing cytokines from
human cells of the different BCC species is also likely determined by the
lipid A acylation pattern. Elegant work on E. coli lipid A59 demonstrates the
highest biological activity correlates with the higher degrees of acylation
degree and to the asymmetrical distribution of the fatty acid alkyl chains
with respect to the two glucosamine residues. Collectively these factors
leads to conformational changes in the lipid A, that are presumed to affect
PAMP receptor engagement. These lipid A changes are closely related to
cytokine induction from host immune cells. The activation of host innate
immunity system is, in all likelihood, elicited by natural blends of variously
acylated bacterial lipid A species99; in case of BCC, clinically pathogenic
strains are likely to express more heavily acylated forms e.g. penta-acyl
lipid A as compared to tetra-acylated forms, and this may be one of the
reasons of the elevated response observed in the colonized organisms.46
Similarly, Porphyromonas gingivalis a Gram negative oral pathogen pro-
duces a LPS with lipid A structural variants differing in the acylation
pattern, with penta-acylated and tetra-acylated forms, that can induce
Carbohydr. Chem., 2012, 38, 13–39 | 27

diametrically opposed effects on the immune system, i.e., by altering the
relative amount of the two lipid A structures, P. gingivalis modulates innate
host response.100
5 General structural features of the core region of BCC LPS

To date there are few studies published of BCC LPS core region structure.
This is surprising as the core region of BCC LPS plays a key role in con-
ferring physico-chemical and biological properties to the outer membrane.
The limited data set means it is impossible to form definitive viewpoint on
the chemical composition and structure of all BCC species.46 A unifying and
characterising feature of BCC strains tested so far is the presence of the
disaccharide D-glycero-D-talo-oct-2-ulosonic acid-(2-4)-3-deoxy-D-manno-
oct-2-ulosonic acid, (Ko-(2-8)-Kdo) in the inner core of BCC LPS.46
Commonly though not universally in BCC strains the Ko can be also gly-
cosylated at position O-8 by a Ara4N residue, therefore up to, three Ara4N
residues can be present within the lipid A-inner core region of BCC LPS.
Within the outer core, heptoses are present, which can be in L,D or the D,D
configuration. To date, no phosphate or other anionic residues have yet been
identified within the outer core of the tested BCC strains.46 The presence of a
complete BCC LPS inner core oligosaccharide is required for in vitro and
in vivo bacterial survival.101 The hldA and hldD genes in B. cenocepacia ET-12
strain K56-2, allow smooth LPS expression and knockdown of this reduces
bacterial persistence. These genes encode enzymes directing the modification
of heptose sugars prior to incorporation into the LPS core oligosaccharide.
A double knockdown mutant, defective in both hldA and hldD expression,
confirmed that both these B. cenocepacia genes were required for synthesis of
a complete LPS core oligosaccharide. The LPS produced by hldA /hldD
mutant strains consisted of a short lipid A-core oligosaccharide devoid of the
O-antigen. Limited hldA /hldD bacterial persistence was noted in a rat
lung infection model in vivo. This may be due to increases sensitivity to
antimicrobial peptides polymyxin B and human neutrophil peptide 1 (HNP-1)
in vitro.101 In contrast, another B. cenocepacia mutant strain that produced
complete lipid A-core oligosaccharide but lacked polymeric O-antigen was
not sensitive to polymyxin B or melittin.101 These data are confirmatory to the
early data showing that rough LPS bearing ‘‘Pseudomonas cepacia’’ isolates
were serum sensitive; collectively these confirm the important virulence role
for core oligosaccharides. Most recent data showed that it is possible to
obtain heptose-less LPS strains of B. cenocepacia with high resistance to
polymyxin B, and suggests that this may occur through LPS-independent
changes that stabilize the outer membrane in some way.102
Here we report all available data on the limited number of BCC LPS core
region structure published.

In B. cepacia genomovar I an incomplete core structure has been shown in
which core oligosaccharide, a heptosyl trisaccharide, is present and directly
linked to O-5 of Kdo, that in turn bears at O-4 the disaccharide Ara4N-
(1-8)-Ko as explained above. The outer core also includes glucose and
28 | Carbohydr. Chem., 2012, 38, 13–39

α-L,D-Hep -(1 →7)-α-L,D-Hep -(1 → 3)-α-L,D-Hep -(1 →5)-α-D-Kdo
2 4 4
↓ ↓ ↓
1 1 1
α-L-Rha β-D-Glc α-D-Ko
Fig. 4 Core oligosaccharide from B. cepacia.
β-D-QuiNAc-(1 →3)-α-L-Rha-(1 →2)-α -L,D-Hep-(1 →3)-α-L,D-Hep-(1 →5)-α-D-Kdo
3 4 4
↓ ↓ ↓
1 1 2
α-L-Rha β-D-Glc α-D-Ko
3 6
↓ ↓
1 1
β-D-Glc-(1 → 3)-α-D-GalNAc-(1 →3)-β-D-GalNAc α-L,D-Hep
Fig. 5 Core oligosaccharide from B. multivorans.
rhamnose. The substitution of Ara4N in the LPS of B. cepacia genomovar I

to date has been found to always occurs on the Ko residue linked to the Kdo
located in the inner core region103 (Fig. 4).
5.2 Burkholderia multivorans genomovar II

B. multivorans LPS core region consists of a much longer and complicate
oligosaccharide in which the inner part matches the structure of B. cepacia
genomovar I whilst the outer consists of a branched oligosaccharide with
residues present in non-stoichiometric amount (Fig. 5) such as the presence
of a [a-D-Glc-(1-3)-a-D-GalNAc-(1-3)-a-D-GalNAc] trisaccharide sitting
on a rhamnose residue.49

The core region of B. cenocepacia has slight differences to those of
B. multivorans above. It consists of a branched oligosaccharide where the
second heptose residue is a 3,7-Hep that carries at O-7 a a-Hep-(1-7)-a-Hep-
(1)- disaccharide.93,94 The outer core region includes galactose, glucose, a
further heptose, that is infrequently found as a component of the outer core
of LOSs (Fig. 6) and, interestingly, a non-stoichiometric terminal dis-
accharide composed by quinovosamine and rhamnose.93 The hypothesis
that this latter disaccharide is a remnant of the O-polysaccharide chain
cannot be ruled out. Furthermore it could potentially be a first repeating
unit of the O-chain, thus giving rise to a semi-rough type LPS.
5.4 Burkholderia pyrrocinia genomovar IX

In a clinical B. pyrrocinia isolate, a rough chemotype LPS was found.
Together with the classical BCC lipid A it possessed a complex oligo-
saccharide in which the Kdo unit was substituted at O-5 by the
Carbohydr. Chem., 2012, 38, 13–39 | 29

α-L,D-Hep-(1→7)-α-L,D-Hep-(1→3)-α-L,D-Hep-(1→5)-α-D-Kdo
3 4 4
↑ ↑ ↑
1 1 2
β-D-QuiNAc-(1→7)-α-L,D-Hep-(1→2)-α-D-Glc β-D-Glc α-D-Ko
3
↑
1
α-L-Rha
4
↑
1
β-D-GalNAc
Fig. 6 Core oligosaccharide from B. cenocepacia.
pentasaccharide a-Hep-(1-7)-[a-Rha-(1-2)]-a-Hep-(1-3)-[b-Glc-(1-4)]-a-Hep-(1-,
a saccharide sequence common to the LOS found in B. cepacia.46 As sub-
stantial difference, the last heptose residue of this portion was however
further substituted by a heptose trisaccharide, giving rise to a heptan pen-
tasaccharide that is the non reducing part of the outer core.46
As for other non BCC core oligosaccharides of Burkholderia LPS, the
only complete data are deriving from our own studies on B. caryophylli
LPS.104 Most strikingly we identified for the first time, that this core region
possesses a unique two L-a-D-Hepp-(1-5)-a-Kdo-2- moieties, one of which is
linked to lipid A with the other linked to Kdo.104
6 O-chain of BCC LPS

There exist several O-antigen structures isolated and characterized from
BCC strains. It is known that BCC species produce many structurally dis-
tinct LPS O-antigens.79,105 In fact, 16 O-antigen serotypes have been iden-
tified among BCC isolates recovered from CF patients.79,106,107 These
serotypes do not correlate with BCC species.108,109 As a consequence,
O-antigen serotype cannot be used to determine the specific BCC species
causing an infection. Nevertheless, all the available data show that the
O-antigens belonging to BCC are rather simple structures with few sugars
constituting the repeating unit, very often containing 6-deoxy sugar (typi-
cally L-rhamnose), and, in the case of the trisaccharide repeating unit, with
at least one sugar present twice. From the immunological point of view, the
presence of rhamnose residues in the O-chain of several Burkholderia LPSs
has been deemed to play a key role in the bacterial interaction with host-
recognition systems, that is, innate plant immunity.110
Several studies elucidated the presence of different O-chain poly-
saccharides within the same BCC strain (B. multivorans, B. vietnamensis)79
in which is unusual in Gram-negative bacteria. Furthermore it has been
demonstrated that bacteria can adopt a strategy during host invasion in
which they undergo ‘‘phase variation’’ involving structural variations in
antigenic determinants;111 thus this heterogeneity in O-antigen structure
within the same strain could be conceived as a particular kind of phase
variation that the bacteria use to augment immunological specificity of the
30 | Carbohydr. Chem., 2012, 38, 13–39

antigenic portion of its LPS, overall contributing to the variety of antigenic
types between BCC species and strains. Analogously, LPS O-acetylation
helps Haemophilus influenza during infection in the human respiratory tract,
facilitating resistance to host immune clearance mechanisms.
As the O-antigen is the most prominent and immunogenic component of
the bacterial cell surface, the structure and composition of the O-antigen
can be used as the basis of a serotype.
We are aware of few studies determining the relationship between BCC
infection in humans and serotype. Rabkin109 et al. isolated BCC from
patients with hospital-acquired infections and the environmental source of
these infections. The serotype of the strains was determined, but, as men-
tioned above, no apparent correlation between the serotype and the source or
clinical outcome was observed.109 More recently, the genomovar status of the
strains used in the prior study of Rabkin et al.109 was determined: strains
from multiple genomovars had the same serotype indicating that genomovar
status and serotype did not correlate.108 There is an obvious need to develop
vaccines to prevent BCC infections in susceptible people, particularly those
living with CF. As the O-antigen is a strong immunogenic molecule, it may be
a potential vaccine candidate for BCC, as has been proposed for other Gram-
negative bacteria, including P. aeruginosa.108 However, as there does not
appear to be a correlation between serotype and disease state, any O-antigen-
based vaccine would need to be multivalent. On the other hand, BCC is being
developed for release into the environment for bioremediation or agricultural
uses. If there is concern that the commercially useful BCC strains may induce
disease in susceptible hosts, a vaccine directed to the O antigen of these
strains could be developed and given to high-risk patients.79
Here we reported several to date BCC O-antigen structures elucidated.

The structure of the O-antigen produced by a clinical isolate of Burkholderia
cepacia isolated from a patient with fibrocystic lung disease has determined
as [-3)a-D-Rhap-(1-3)a-D-Rhap-(1-4)-a-D-Galp-(1-]n.112 Also, an
additional O-antigen from a B. cepacia strain varies greatly; the structure
was determined to be a disaccharide repeat of a glucopyranose followed by
a mannopyranose.113 The O-antigenic polysaccharides of the lipopoly-
saccharides produced by three isolates of Burkholderia cepacia serogroup
A were analysed and showed the repeating trisaccharide unit [-3)-a-D-
GalpNAc-(1-3)-b-D-GalpNAc-(1-4)-a-L-Rhap-(1-]n.114 Moreover, an
isolate of serogroup O7 strain ATCC 17759 has been showed to possess a
the disaccharidic unit [-4)-a-D-Glcp-(1-3)-a-D-Manp-(1-].115
6.2 Burkholderia multivorans genomovar II

Burkholderia multivorans LPS possesses a mixture of two O-chains with
different repeating units: [-2-a-D-Man-(1-2)-a-D-Rha-(1-3)-a-D-Man-1-]n
(species X, 60%) and [-2-a-D-Man-(1-2)-a-D-Aco-(1-3)-a-D-Rha1-]n
(species Y, 40%).94 As mentioned above, the two trisaccharide repeating
units are both presenting a-manno-configured 6-deoxy sugars; in both cases
a sugar is repeated twice (a-mannose in the first polysaccharide and
a-rhamnose in the second type). Some other important and unusual features
Carbohydr. Chem., 2012, 38, 13–39 | 31

relevant peculiarities can be found in these structures; they both present D-
configured rhamnose, which within BCC LPS was only previously found in
the O-polysaccharide of B. cepacia.116 Furthermore, in species X the 2 sub-
stituted a-Rha is stoichiometrically substituted by a methyl group at O-3
(acofriose). The presence of non-carbohydrate appendages, such as a methyl
group, is common to several antigens isolated from BCC strains, often pre-
senting O-acyl decorations (for example, in strain CIP8240 there is a non-
stoichiometric acylation at O-6 of a-Galf, in CCUG7246 at O-3 of a-Man, in
ATCC 17 759 at O-2 of a-Man) although the O-methyl substitution has been
found for the first time to be decorating an O-polysaccharide from Bur-
kholderia LPS.

B. cenocepacia strains have been shown to have a [a-D-GalNAc-(1-3)-a-D-
GalNAc-(1-4)-a-L-Rha]n trisaccharide sitting on a quinovosamine residue.
Importantly the O polysaccharide has been demonstrated to contribute to
immune responses or immune evasion.117,118 Medicinal chemistry approa-
ches have led to the organic synthesis of the trisaccharide mimicking
the repeating unit of an O-polysaccharide portion of LPS present in a
clinical isolate of Burkholderia cepacia. This may offer a possible poly-
saccharide based vaccine. However these approaches need to be extended
to B. cenocepacia and B. multivorans as the most clinically prevalent strains
in CF.
6.4 Burkholderia stabilis genomovar IV

Burkholderia stabilis strain CIP 8235 have two different O-chain structures
characterised by a disaccharidic repeating unit: a major component is [-4)-
a-D-Glcp-(1-3)-a-L-GlcpNAc-(1-] and a minor component is [-4)-a-D-
Glcp-(1-3)-a-L-Rhap-(1-].119,120
6.5 Burkholderia vietnamiensis genomovar V

Burkholderia vietnamiensis strain LMG 6998 possesses the linear tri-
saccharide repeating-unit [-3)-a-L-Rhap-(1-3)-b-D-Manp-(1-4)-a-D-
Manp-(1-]n76 where the 4-substituted mannose residue carries an O-acetyl
group at position 3. Whilst lipopolisaccharide isolated from strain LMG
6999 of Burkholderia vietnamiensis is characterised by the presence of two
polymeric fractions based on the following trisaccharide repeating units: I:
[-3)-a-D-Galp-(1-3)-b-D-Galp-(1-3)-b-D-GalpNAc-(1-]n; II: [-3)-a-
121
D-GalpNAc-(1-3)-b-D-GalpNAc-(1-4)-a-L-Rhap-(1-]n.
6.6 Burkholderia anthina genomovar VIII

The structure of the O-antigen from a strain of B. anthina is [-3)a-L-Rhap-
(1-2)a-L-Rhap-(1-2)-a-D-Galp-(1-]n.122 This structure is similar to an
O-antigen that was identified in a B. cepacia clinical isolate: [-3)a-D-Rhap-
(1-3)-a-D-Rhap-(1-4)-a-D-Galp-(1-]n.
Searching for a rationale in the aforementioned heterogeneity of the O-
antigen subunit structures from BCC, there are three things to notice. First,
many of the O-antigen repeating unit are simple structures, having only two
or three sugar residues per subunit. Those subunits with three sugars have at
32 | Carbohydr. Chem., 2012, 38, 13–39

least one sugar residue repeated: for example, the O antigen of the serotype
K strain, IMV 3181, is a polyrhamnose. Second, there are some unusual
sugar residues found in the O-antigen structures from BCC; for example,
L,D-heptose is usually found only in the core oligosaccharide of other
Gram-negative bacteria, and the D forms of rhamnose and fucose are found
in O-antigen subunits of BCC and other bacteria of the Pseudomonadaceae
family as opposed to the L forms of these sugars.121,119,123 Third, there are
two O-antigen subunit structures of BCC that are in common with
P. aeruginosa O-antigen subunits: the O-antigen subunits of BCC serotype
O3 and O5 are identical to the O-antigen subunits of P. aeruginosa ser-
ogroup O15 and O17, respectively.124
7 Conclusion
Cystic Fibrosis provides a unique clinical challenge. The complex envir-
onment within the lung and the changing milieu is associated with strong
local bacterial environmental pressures that subsequently alter many
aspects of the bacteria virulence. The inflammatory processes within the
lung can be traced at least in part to the biological activities of the lipo-
polysaccharide moieties. Specific changes in LPS occur in two key pathogen
groups Pseudomonas aeruginosa and Burkholderia spp. To date these
changes seem specific to CF and have not been seen in other human
infections. ‘‘CF specific’’ changes in LPS appear to occur relatively rapidly
after these organisms are acquired in CF and are maintained in vitro. These
confer an advantage for persistence of the bacteria but the mechanisms
involved may also offer opportunities for manipulation therapeutically.
Novel data arising from lung transplantation strains suggest that environ-
mental changes in vivo are associated with further changes in LPS and lipid
A from the chronic CF type LPS. The potential to ‘‘revert’’ CF specific
changes in LPS composition is interesting though the consequences need a
comprehensive analysis. Unfortunately the lack of good animal models and
the host specific interaction between LPS and host TLR demonstrated for
Pseudomonas makes such studies problematic. The use of clinical strains in
research from both early acquisition and later stages of CF lung disease is
mandatory to reflect the spectrum of disease. Environmental Pseudomonas
strains have different virulence and LPS characteristics based on the
available data. Hence studies for vaccine development or therapeutic
manipulation of LPS biosynthesis must avoid the over-representation of
environmental strains as compared to clinically derived strains. Concerted
efforts across traditional scientific boundaries of chemistry, microbiology,
immunology and medicine will be required to make major advances in our
understanding of these areas. Notably several collaborative have formed to
this end e.g. http://www.cost.eu/domains_actions/bmbs/Actions/BM1003.
Future studies should carefully consider the growth conditions used in vitro
and where possible consideration of testing more than one strain from a
patient may yield more generalisable data. The ability to access a standar-
dised panel of CF Pseudomonas strains such as is available for the inter-
national Burkholderia reference panel may well allow more rapid advances
to occur.
Carbohydr. Chem., 2012, 38, 13–39 | 33

Acknowledegments
We all acknowledge Cost Action BM1003 ‘‘Microbial cell surface deter-
minants of virulence as targets for new therapeutics in Cystic Fibrosis’’.
ADS is a Higher Education Funding Council for England (HEFCE) Senior
lecturer. We acknowledge funding from the Newcastle Hospitals Special
Trustees (ADS), The Italian Cystic Fibrosis Foundation (FFC; grant 09/16
(ADS and AS), FFC 11#2010 with the contribution of Pastificio Giovanni
Rana s.p.a. to AM). A.M, A.S.,and R.L acknowledge PRIN MIUR 2008/
2009 (A.M. and R.L.).
References
1 F. Ratjen and H. Grasemann, Current Pharmaceutical Design, 2012.
2 C. S. Halliburton, D. M. Mannino and R. S. Olney, Archives of Pediatrics &
Adolescent Medicine, 1996, 150, 1181–1185.
3 T. G. Liou, F. R. Adler, S. C. Fitzsimmons, B. C. Cahill, J. R. Hibbs and B. C.
Marshall, American Journal of Epidemiology, 2001, 153, 345–352.
4 M. I. Gomez and A. Prince, Curr. Opin. Pharmacol., 2007, 7, 244–251.
5 E. Mahenthiralingam, T. A. Urban and J. B. Goldberg, Nat. Rev. Microbiol.,
2005, 3, 144–156.
6 T. Coenye, J. Goris, T. Spilker, P. Vandamme and J. J. LiPuma, Journal of
Clinical Microbiology, 2002, 40, 2062–2069.
7 J. S. Brooke, Clinical Microbiology Reviews, 2012, 25, 2–41.
8 J. Emerson, S. McNamara, A. M. Buccat, K. Worrell and J. L. Burns,
Pediatric Pulmonology, 45, 363–370.
9 L. Spicuzza, C. Sciuto, G. Vitaliti, G. Di Dio, S. Leonardi and M. La Rosa,
Eur. J. Clin. Microbiol. Infect. Dis., 2009, 28, 191–195.
10 C. Ronne Hansen, T. Pressler, N. Hoiby and M. Gormsen, J. Cyst. Fibros.,
2006, 5, 245–251.
11 J. R. Govan and V. Deretic, Microbiological Reviews, 1996, 60, 539–574.
12 S. M. Bianchi, L. R. Prince, K. McPhillips, L. Allen, H. M. Marriott, G. W.
Taylor, P. G. Hellewell, I. Sabroe, D. H. Dockrell, P. W. Henson and M. K.
Whyte, American Journal of Respiratory and Critical Care Medicine, 2008, 177,
35–43.
13 M. Rosenfeld, R. L. Gibson, S. McNamara, J. Emerson, J. L. Burns, R.
Castile, P. Hiatt, K. McCoy, C. B. Wilson, A. Inglis, A. Smith, T. R. Martin
and B. W. Ramsey, Pediatric Pulmonology, 2001, 32, 356–366.
14 E. Kerem, M. Corey, R. Gold and H. Levison, Journal of Pediatrics, 1990, 116,
714–719.
15 N. Lechtzin, M. John, R. Irizarry, C. Merlo, G. B. Diette and M. P. Boyle,
Respiration; International Review of Thoracic Diseases, 2006, 73, 27–33.
16 N. J. Simmonds, S. J. Macneill, P. Cullinan and M. E. Hodson, Eur. Respir. J.,
2011, 36, 1277–1283.
17 K. Cheng, R. L. Smyth, J. R. Govan, C. Doherty, C. Winstanley, N. Denning,
D. P. Heaf, H. van Saene and C. A. Hart, Lancet, 1996, 348, 639–642.
18 A. Ashish, M. Shaw, C. Winstanley, M. J. Ledson and M. J. Walshaw, J. Cyst.
Fibros.
19 M. Al-Aloul, J. Crawley, C. Winstanley, C. A. Hart, M. J. Ledson and M. J.
Walshaw, Thorax, 2004, 59, 334–336.
20 E. Mahenthiralingam, B. Bischof, S. K. Byrne, C. Radomski, J. E. Davies, Y.
Av-Gay and P. Vandamme, Journal of Clinical Microbiology, 2000, 38, 3165–3173.
21 L. Saiman, Current Opinion in Infectious Diseases, 2012, 24, 390–395.
34 | Carbohydr. Chem., 2012, 38, 13–39

22 L. Eberl and B. Tummler, Int. J. Med. Microbiol., 2004, 294, 123–131.
23 A. R. Hauser, M. Jain, M. Bar-Meir and S. A. McColley, Clin. Microbiol.
Rev., 2011, 24, 29–70.
24 G. Doring, I. G. Parameswaran and T. F. Murphy, FEMS Microbiology
Reviews.
25 E. E. Smith, D. G. Buckley, Z. Wu, C. Saenphimmachak, L. R. Hoffman, D.
A. D’Argenio, S. I. Miller, B. W. Ramsey, D. P. Speert, S. M. Moskowitz, J. L.
Burns, R. Kaul and M. V. Olson, Proceedings of the National Academy of
Sciences of the United States of America, 2006, 103, 8487–8492.
26 D. J. Waine, D. Honeybourne, E. G. Smith, J. L. Whitehouse and C. G.
Dowson, Journal of Clinical Microbiology, 2008, 46, 3491–3493.
27 M. T. Rybtke, P. O. Jensen, N. Hoiby, M. Givskov, T. Tolker-Nielsen and T.
Bjarnsholt, Inflammation & Allergy Drug Targets, 10, 141–157.
28 N. Hoiby, O. Ciofu and T. Bjarnsholt, Future Microbiology, 5, 1663–1674.
29 S. Elias and E. Banin, FEMS Microbiology Reviews, 2012.
30 M. J. Hunter, K. J. Treharne, A. K. Winter, D. M. Cassidy, S. Land and
A. Mehta, PloS One, 2010, 5, e11598.
31 L. S. Nixon, B. Yung, S. C. Bell, J. S. Elborn and D. J. Shale, American
Journal of Respiratory & Critical Care Medicine, 1998, 157, 1764–1769.
32 P. S. Salva, N. A. Doyle, L. Graham, H. Eigen and C. M. Doerschuk,
Pediatric Pulmonology, 1996, 21, 11–19.
33 E. Raoust, V. Balloy, I. Garcia-Verdugo, L. Touqui, R. Ramphal and M.
Chignard, PloS One, 2009, 4, e7259.
34 S. M. Rowe, P. L. Jackson, G. Liu, M. Hardison, A. Livraghi, G. M. Solo-
mon, D. B. McQuaid, B. D. Noerager, A. Gaggar, J. P. Clancy, W. O’Neal, E.
J. Sorscher, E. Abraham and J. E. Blalock, American Journal of Respiratory
and Critical Care Medicine, 2008, 178, 822–831.
35 M. Brodlie, M. C. McKean, G. E. Johnson, J. Gray, A. J. Fisher, P. A. Corris,
J. L. Lordan and C. Ward, Am. J. Respir. Crit. Care Med., 2011, 182, 369–375.
36 N. Regamey, L. Tsartsali, T. N. Hilliard, O. Fuchs, H. L. Tan, J. Zhu, Y. S.
Qiu, E. W. Alton, P. K. Jeffery, A. Bush and J. C. Davies, Thorax, 2012, 67,
170–176.
37 S. Voglis, K. Quinn, E. Tullis, M. Liu, M. Henriques, C. Zubrinich, O.
Penuelas, H. Chan, F. Silverman, V. Cherepanov, N. Orzech, A. A. Khine, A.
Cantin, A. S. Slutsky, G. P. Downey and H. Zhang, American Journal of
Respiratory and Critical Care Medicine, 2009, 180, 159–166.
38 D. Hartl, P. Latzin, P. Hordijk, V. Marcos, C. Rudolph, M. Woischnik, S.
Krauss-Etschmann, B. Koller, D. Reinhardt, A. A. Roscher, D. Roos and M.
Griese, Nature Medicine, 2007, 13, 1423–1430.
39 G. B. Rogers, C. A. Hart, J. R. Mason, M. Hughes, M. J. Walshaw and K. D.
Bruce, Journal of Clinical Microbiology, 2003, 41, 3548–3558.
40 G. B. Rogers, M. P. Carroll, D. J. Serisier, P. M. Hockey, G. Jones, V.
Kehagia, G. J. Connett and K. D. Bruce, Journal of Clinical Microbiology,
2006, 44, 2601–2604.
41 M. S. Doud, M. Light, G. Gonzalez, G. Narasimhan and K. Mathee, Human
Genomics, 2010, 4, 147–169.
42 M. M. Tunney, T. R. Field, T. F. Moriarty, S. Patrick, G. Doering, M. S.
Muhlebach, M. C. Wolfgang, R. Boucher, D. F. Gilpin, A. McDowell and J.
S. Elborn, American Journal of Respiratory and Critical Care Medicine, 2008,
177, 995–1001.
43 T. R. Field, A. White, J. S. Elborn and M. M. Tunney, Eur. J. Clin. Microbiol.
Infect. Dis., 2005, 24, 677–687.
44 N. Bagge, O. Ciofu, L. T. Skovgaard and N. Hoiby, Apmis, 2000, 108, 589–600.
Carbohydr. Chem., 2012, 38, 13–39 | 35

45 A. Silipo, G. Erbs, T. Shinya, J. M. Dow, M. Parrilli, R. Lanzetta, N. Shibuya,
M. A. Newman and A. Molinaro, Glycobiology, 2010, 20, 406–419.
46 A. De Soyza, A. Silipo, R. Lanzetta, J. R. Govan and A. Molinaro, Innate
Immunity, 2008, 14, 127–144.
47 C. R. Raetz and C. Whitfield, Annu. Rev. Biochem., 2002, 71, 635–700.
48 N. L. Que, S. Lin, R. J. Cotter and C. R. Raetz, The Journal of Biological
Chemistry, 2000, 275, 28006–28016.
49 T. Ierano, A. Silipo, L. Sturiale, D. Garozzo, H. Brookes, C. M. Khan, C.
Bryant, F. K. Gould, P. A. Corris, R. Lanzetta, M. Parrilli, A. De Soyza and
A. Molinaro, Glycobiology, 2008, 18, 871–881.
50 S. Sforza, A. Silipo, A. Molinaro, R. Marchelli, M. Parrilli and R. Lanzetta,
J. Mass Spectrom., 2004, 39, 378–383.
51 L. Sturiale, D. Garozzo, A. Silipo, R. Lanzetta, M. Parrilli and A. Molinaro,
Rapid Commun. Mass Spectrom., 2005, 19, 1829–34.
52 S. Hase and E.T. Rietschel, Eur. J. Biochem., 1976, 63, 101–7.
53 Y. A. Knirel, J. H. Helbig and U. Zähringer, Carbohydr. Res., 1996, 283, 129–39.
54 L. Brecker, Chem. Phys. Lipids, 2003, 125, 27–39.
55 A. Silipo, C. De Castro, R. Lanzetta, A. Molinaro and M. Parrilli, Glyco-
biology, 2004, 14, 805–15.
56 Y. A. Knirel, N. A. Kocharova, S. O. Hynes, G. Widmalm, L. P. Andersen,
P. E. Jansson and A. P. Moran, European Journal of Biochemistry/FEB, 1999,
266, 123–131.
57 O. Holst, Methods Mol. Biol., 2000, 145, 345–53.
58 A. Silipo, R. Lanzetta, M. Parrilli, L. Sturiale, D. Garozzo, E. L. Nazarenko,
R. P. Gorshkova, E. P. Ivanova and A. Molinaro, Carbohydrate Research,
2005, 340, 1475–1482.
59 P. Rallabhandi, A. Awomoyi, K. E. Thomas, A. Phalipon, Y. Fujimoto, K.
Fukase, S. Kusumoto, N. Qureshi, M. B. Sztein and S. N. Vogel, J. Immunol.,
2008, 180, 1139–1147.
60 S. M. Moskowitz and R. K. Ernst, Sub-cellular Biochemistry, 2010, 53, 241–253.
61 J. S. Lam, V. L. Taylor, S. T. Islam, Y. Hao and D. Kocincova, Frontiers in
Microbiology, 2011, 2, 118.
62 R. K. Ernst, A. M. Hajjar, J. H. Tsai, S. M. Moskowitz, C. B. Wilson and S. I.
Miller, Journal of Endotoxin Research, 2003, 9, 395–400.
63 O. V. Bystrova, A. S. Shashkov, N. A. Kocharova, Y. A. Knirel, U. Zahringer
and G. B. Pier, Biochemistry. Biokhimiia, 2003, 68, 918–925.
64 F. Beckmann, H. Moll, K. E. Jäger and U. Zähringer, Carbohydr. Res., 1995,
267, C3–7.
65 Y. A. Knirel, J. H. Helbig and U. Zähringer, Carbohydr. Res., 1996, 283, 129–39.
66 O. V. Bystrova, A. S. Shashkov, N. A. Kocharova, Y. A. Knirel, B. Lindner,
U. Zahringer and G. B. Pier, European Journal of Biochemistry/FEB, 2002,
269, 2194–2203.
67 Y. A. Knirel, O. V. Bystrova, A. S. Shashkov, B. Lindner, N. A. Kocharova, S.
N. Senchenkova, H. Moll, U. Zahringer, K. Hatano and G. B. Pier, European
Journal of Biochemistry/FEBS, 2001, 268, 4708–4719.
68 A. G. Walsh, M. J. Matewish, L. L. Burrows, M.A. Monteiro, M. B. Perry and
J. S. Lam, Mol. Microbiol., 2000, 35, 718–27.
69 O. V. Bystrova, Y. A. Knirel, B. Lindner, N. A. Kocharova, A. N. Kondakova,
U. Zahringer and G. B. Pier, FEMS Immunology and Medical Microbiology,
2006, 46, 85–99.
70 D. J. Evans, G. B. Pier, M. J. Coyne and J. B. Goldberg, Mol. Microbiol.,
1994, 13, 427–34.
71 Y. A. Knirel, Critical Reviews in Microbiology, 1990, 17, 273–304.
36 | Carbohydr. Chem., 2012, 38, 13–39

72 S. I. Miller, R. K. Ernst and M. W. Bader, Nat. Rev. Microbiol., 2005, 3, 36–46.
73 M. T. Albrecht, W. Wang, O. Shamova, R. I. Lehrer and N. L. Schiller,
Respir. Res., 2002, 3, 18.
74 J. B. McPhee, S. Lewenza and R. E. Hancock, Molecular Microbiology, 2003,
50, 205–217.
75 J. B. McPhee, M. Bains, G. Winsor, S. Lewenza, A. Kwasnicka, M. D. Brazas,
F. S. Brinkman and R. E. Hancock, Journal of Bacteriology, 2006, 188,
3995–4006.
76 E. L. Macfarlane, A. Kwasnicka, M. M. Ochs and R. E. Hancock, Molecular
Microbiology, 1999, 34, 305–316.
77 C. Cigana, L. Curcuru, M. R. Leone, T. Ierano, N. I. Lore, I. Bianconi, A.
Silipo, F. Cozzolino, R. Lanzetta, A. Molinaro, M. L. Bernardini and A.
Bragonzi, PloS one, 2009, 4, e8439.
78 A. M. Hajjar, R. K. Ernst, J. H. Tsai, C. B. Wilson and S. I. Miller, Nature
Immunology, 2002, 3, 354–359.
79 A. D. Vinion-Dubiel and J. B. Goldberg, Journal of Endotoxin Research, 2003,
9, 201–213.
80 D. P. Speert, Paediatric Respiratory Reviews, 2002, 3, 230–235.
81 A. Agodi, E. Mahenthiralingam, M. Barchitta, V. Giannino, A. Sciacca and S.
Stefani, Journal of Clinical Microbiology, 2001, 39, 2891–2896.
82 J. M. van den Berg, E. van Koppen, A. Ahlin, B. H. Belohradsky, E. Berna-
towska, L. Corbeel, T. Espanol, A. Fischer, M. Kurenko-Deptuch, R. Mouy,
T. Petropoulou, J. Roesler, R. Seger, M. J. Stasia, N. H. Valerius, R. S.
Weening, B. Wolach, D. Roos and T. W. Kuijpers, PloS one, 2009, 4, e5234.
83 A. De Soyza, K. Morris, A. McDowell, C. Doherty, L. Archer, J. Perry, J. R.
Govan, P. A. Corris and K. Gould, Thorax, 2004, 59, 526–528.
84 A. De Soyza and P. A. Corris, J. Heart Lung Transplant., 2003, 22, 954–958.
85 D. Hadjiliadis, M. P. Steele, C. Chaparro, L. G. Singer, T. K. Waddell, M. A.
Hutcheon, R. D. Davis, D. E. Tullis, S. M. Palmer and S. Keshavjee, J. Heart
Lung Transplant., 2007, 26, 834–838.
86 B. D. Alexander, E. W. Petzold, L. B. Reller, S. M. Palmer, R. D. Davis, C. W.
Woods and J. J. Lipuma, Am. J. Transplant., 2008, 8, 1025–1030.
87 P. M. Hopkins, T. J. Kidd, C. Coulter, I. H. Feather, P. Derrington and S. C.
Bell, Am. J. Respir. Crit. Care Med., 2009, 179, 257–258.
88 S. Murray, J. Charbeneau, B. C. Marshall and J. J. LiPuma, Am. J. Respir.
Crit. Care Med., 2008, 178, 363–371.
89 A. De Soyza, G. Meachery, K. L. Hester, A. Nicholson, G. Parry, K.
Tocewicz, T. Pillay, S. Clark, J. L. Lordan, S. Schueler, A. J. Fisher, J. H.
Dark, F. K. Gould and P. A. Corris, J. Heart Lung Transplant., 2010.
90 E. Mahenthiralingam, T. Coenye, J. W. Chung, D. P. Speert, J. R. Govan, P.
Taylor and P. Vandamme, Journal of Clinical Microbiology, 2000, 38, 910–913.
91 A. De Soyza, C. D. Ellis, C. M. Khan, P. A. Corris and R. Demarco de
Hormaeche, American Journal of Respiratory and Critical Care Medicine,
2004, 170, 70–77.
92 A. Silipo, A. Molinaro, T. Ierano, A. De Soyza, L. Sturiale, D. Garozzo, C.
Aldridge, P. A. Corris, C. M. Khan, R. Lanzetta and M. Parrilli, Chemistry,
2007, 13, 3501–3511.
93 A. De Soyza, C. D. Ellis, C. M. A. Anjam, P. A. Corris and R. DeMarco De
Hormaeche, American Journal of Respiratory & Critical Care Medicine, 2004,
170, 70–77.
94 T. Ierano, A. Silipo, L. Sturiale, D. Garozzo, C. Bryant, R. Lanzetta, M.
Parrilli, C. Aldridge, F. K. Gould, P. A. Corris, C. M. Khan, A. De Soyza and
A. Molinaro, Glycobiology, 2009, 19, 1214–1223.
Carbohydr. Chem., 2012, 38, 13–39 | 37

95 H. Iihara, T. Niwa, M. M. Shah, P. H. Nhung, S. X. Song, M. Hayashi, K.
Ohkusa, Y. Itoh, S. Makino and T. Ezaki, Jpn. J. Infect Dis., 2007, 60, 230–4.
96 A. Silipo, A. Molinaro, P. Cescutti, E. Bedini, R. Rizzo, M. Parrilli and R.
Lanzetta, Glycobiology, 2005, 15, 561–570.
97 A. Molinaro, C. De Castro, R. Lanzetta, A. Evidente, M. Parrilli and O.
Holst, The Journal of Biological Chemistry, 2002, 277, 10058–10063.
98 H. Shimomura, M. Matsuura, S. Saito, Y. Hirai, Y. Isshiki and K. Kawahara,
Infection and Immunity, 2003, 71, 5225–5230.
99 G. Erbs, T. T. Jensen, A. Silipo, W. Grant, J. M. Dow, A. Molinaro, M.
Parrilli and M. A. Newman, Microbes and Infection/Institut Pasteur, 2008, 10,
571–574.
100 R. A. Reife, S. R. Coats, M. Al-Qutub, D. M. Dixon, P. A. Braham, R. J.
Billharz, W. N. Howald and R. P. Darveau, Cellular Microbiology, 2006, 8,
857–868.
101 S. A. Loutet, R. S. Flannagan, C. Kooi, P. A. Sokol and M. A. Valvano,
Journal of Bacteriology, 2006, 188, 2073–2080.
102 S. A. Loutet, F. Di Lorenzo, C. Clarke, A. Molinaro and M. A. Valvano,
BMC genomics, 2011, 12, 472.
103 Y. Isshiki, U. Zahringer and K. Kawahara, Carbohydrate Research, 2003, 338,
2659–2666.
104 C. De Castro, A. Molinaro, R. Lanzetta, O. Holst and M. Parrilli, Carbohy-
drate Research, 2005, 340, 1802–1807.
105 M. N. Burtnick, P. J. Brett and D. E. Woods, Journal of Bacteriology, 2002,
184, 849–852.
106 Y. Nakamura, S. Hyodo, E. Chonan, S. Shigeta and E. Yabuuchi, Journal of
Clinical Microbiology, 1986, 24, 152–154.
107 A. Soldatkina, A. Knirel’ Iu, N. V. Tanatar and I. Zakharova, Mikrobiolo-
gicheskii zhurnal, 1989, 51, 32–38.
108 D. T. Kenna, V. A. Barcus, R. J. Langley, P. Vandamme and J. R. Govan,
FEMS Immunology Medical Microbiology, 2003, 35, 87–92.
109 C. S. Rabkin, W. R. Jarvis, R. L. Anderson, J. Govan, J. Klinger, J. LiPuma,
W. J. Martone, H. Monteil, C. Richard and S. Shigeta, et al., Rev. Infect. Dis.,
1989, 11, 600–607.
110 K. A. Mattos, A. R. Todeschini, N. Heise, C. Jones, J. O. Previato and L.
Mendonca-Previato, Glycobiology, 2005, 15, 313–321.
111 M. Lukáčová, I. Barák and J. Kazár, Clin. Microbiol. Infect., 2008, 14, 200–6.
112 S. Cerantola, A. Lemassu-Jacquier and H. Montrozier, European Journal of
Biochemistry, 1999, 260, 373–383.
113 A. D. Cox and S. G. Wilkinson, Carbohydrate research, 1990, 195, 295–301.
114 L. M. Beynon and M. B. Perry, Biochemistry and Cell Biology=Bichimie et
biologie Cellulaire, 1993, 71, 417–420.
115 A. Knirel’ Iu, N. V. Tanatar, M. A Soldatkina, A. S. Shashkov and I.
Zakharova, Bioorganicheskaia Khimiia, 1988, 14, 77–81.
116 S. Cerantola, A. Lemassu-Jacquier and H. Montrozier, European Journal of
Biochemistry/FEB, 1999, 260, 373–383.
117 S. Kotrange, B. Kopp, A. Akhter, D. Abdelaziz, A. Abu Khweek, K. Caution,
B. Abdulrahman, M. D. Wewers, K. McCoy, C. Marsh, S. A. Loutet, X.
Ortega, M. A. Valvano and A. O. Amer, Journal of Leukocyte Biology, 2011,
89, 481–488.
118 M. S. Saldias, X. Ortega and M. A. Valvano, Journal of Medical Microbiology,
2009, 58, 1542–1548.
119 N. A. Paramonov, A. S. Shashkov, A. Knirel Iu, M. A Soldatkina and I.
Zakharova, Bioorganicheskaia Khimiia, 1994, 20, 984–993.
38 | Carbohydr. Chem., 2012, 38, 13–39

120 A. D. Cox, C. J. Taylor, A. J. Anderson, M. B. Perry and S. G. Wilkinson,
European Journal of Biochemistry/FEB, 1995, 231, 784–789.
121 D. Gaur and S. G. Wilkinson, Carbohydrate Research, 1996, 295, 179–184.
122 S. Carillo, A. Silipo, V. Perino, R. Lanzetta, M. Parrilli and A. Molinaro,
Carbohydrate Research, 2009, 344, 1697–1700.
123 A. Knirel’ Iu, M. A. Soldatkina, A. S. Shashkov, N. V. Tanamar and N. A.
Paramonov, Bioorganicheskaia Khimiia, 1988, 14, 1678–1683.
124 A. D. Cox and S. G. Wilkinson, Carbohydrate Research, 1989, 195, 123–129.
Carbohydr. Chem., 2012, 38, 13–39 | 39

Synthesis of bacterial carbohydrate surface
structures containing Kdo and glycero-D-
manno-heptose linkages
Stefan Oscarson
DOI: 10.1039/9781849734769-00040
1 Introduction
The inherent complexity of oligosaccharides, due to the large amount
of hydroxyl groups and the stereochemistry of the glycosidic linkage, is
a major drawback in their synthesis. Compared to human structures
microbial ones are even more challenging because of a vast variety of
possible monosaccharide building blocks as well as substituents. In 2000 we
wrote a review on the synthesis of bacterial structures containing two major
synthetic challenges, inner core LPS-structures with the unusual sugars Kdo
and L-glycero-D-manno-heptose and CPS structures built up from repeating
units linked via phosphodiester linkages.1 Herein an update on the progress
in the synthesis of Kdo and glycero-D-manno-heptose containing structures
is presented. Recently this field has been reviewed also by Kosma.2,3
2 Kdo-containing structures and Kdo donors

In spite of the biological importance of Kdo-containing oligosaccharide
structures, the synthetic interest in Kdo is to a large extent focused on the
stereoselective synthesis of the monosaccharide and not in its subsequent
use in oligosaccharide synthesis. Thus, annually a number of synthesis of
Kdo are published, which have been reviewed.4,5 However, most of these,
although conceptually elegant, are multi-step syntheses. For a quick and
reproducible supply of large quantities of Kdo, needed for continued oligo-
saccharide synthesis, the one-step Cornforth procedure,6 originally developed
for synthesis of sialic acid,7 is still most often the method of choice. Starting
from commercial D-arabinose and oxaloacetic acid, a classical condensation
followed by decarboxylation affords Kdo, which usually is isolated as its
crystalline acetylated methyl ester.
Apart from not being commercially available there are other problems
associated with Kdo in oligosaccharide synthesis, the easy formation of the
a,b-unsaturated acid derivative, i.e., a glycal derivative, and the lack of a
stereo-directing group in the 3-position. The former is a problem both in
protective group manipulation and in glycosylation reactions, since the
glycal is formed both under basic and acidic conditions. These problems
(and structural features) are similar to the ones encountered with
neuraminic acid. In spite of the similarities it is still difficult to transfer
experiences between the two sugars, mainly because of the difference in ring
conformation, sialic acid commonly being in the 1C4-conformation with
Centre for Synthesis and Chemical Biology, University College Dublin, Belfield, Dublin 4,
Ireland. E-mail: stefan.oscarson@ucd.ie
40 | Carbohydr. Chem., 2012, 38, 40–60

an equatorial glycosidic a-linkage, whereas Kdo occurs in Nature in the
4
C1-conformation with an axial glycosidic a-linkage.
In the earlier review, a survey on the synthesis of Kdo-containing
saccharides up to 1999 was made.1 Several complex structures had been
produced especially by the Paulsen group from Hamburg, but sometimes
the glycosylation yields were rather low as well as the stereoselectivity. That
far mainly halide donors (especially the bromide) had been used in Kdo
glycosylation reactions.
Kosma and co-workers have continued their synthesis of Chlamydia
structures and analogues thereof. Hence the branched trisaccharide motif
a-Kdo-(2-4)-[a-Kdo-(2-8)-]-a-Kdo of Chlamydophila psittaci was syn-
thesized using acetobromo-Kdo as donor, mercury salts as promoters, and
nitromethane as solvent.8,9 The trisaccharide was synthesized starting either
with the (2-4)-linkage8 or the (2-8)-linkage.9 The Kdo-trisaccharide
was also converted into a bromide donor and used in a coupling to give a
(2-4)-linked Kdo tetrasaccharide derivative (Scheme 1).8 Especially in the
latter glycosylation glycal formation was a major side reaction (66%),
however, a-stereoselectivity was very high throughout the syntheses.
The Kosma group also, very recently, published the synthesis of a
a-Kdo-(2-4)-[b-L-Ara4N-(1-8)-]-a-Kdo structure from Burkholderia and
AcO OAc
AcO
O HO OH
CO2Me
AcO AcO
AcO OAc
O CO2Me AcO
O
O CO2Me
AcO OAc AcO
AcO OAc
OAll AcO
AcO
O HO O
O CO2Me
CO2Me Hg(CN)2 AcO
AcO AcO
MeNO2
59% O CO2Me
O
Br
HgBr2, Hg(CN)2
O R
AcO OAc MeNO2
R = OAll
AcO O
O R = Br
O COOMe HO
AcO OAc AcO O CO2Me
AcO HO
O HO O
CO2Me O
AcO AcO OAll
O CO2Me O
O O
HO
O COOMe
32% (+3% β)
O
OAll
Scheme 1
Carbohydr. Chem., 2012, 38, 40–60 | 41

AcO OAc
AcO
O N3
CO2Me
AcO O
N3 AcO OAc BnO
O Br AcO BnO
BnO HO O
O CO2Me
BnO AcO HO
HO O
HgBr2, Hg(CN)2 O CO2Me
HO O
MeNO2
O CO2Me
HO
19% (+1.4% β) OAll
OAll
Scheme 2
Proteus LPS.10 Here, the branching (2-4)-linked Kdo moiety was

introduced using a Kdo 4,5,7-triol acceptor. As expected the 5-OH was not
involved in glycosylation, but the regioselectivity between the 4- and 7-OH
was a problem. In CH2Cl2 the (2-7)-linked trisaccharide was dominating
over the desired (2-4)-linked product, but in nitromethane this selectivity
was reversed. However, only a 19% yield of the a-(2-4)-linked tri-
saccharide was obtained (together with 1.4% of the b-(2-4)-linked and
5.5% of the a-(2-7)-linked trisaccharide) (Scheme 2).
Kusumoto and coworkers continued their use of isopropylidene acetal
protected Kdo fluoride donors as a solution to the stereoselectivity problem,
the rationale being that the protecting groups force the ring into a boat-like
conformation, which increase the amount of a-side attack and thus
formation of the a-glycoside.11,12 In an elegant synthesis of the complex
complete Re-type LPS structure, the authors showed that a 4,5-O-iso-
propylidene group is the important factor, and also that TBDMS protecting
groups in the 4,5-positions gives the same effect, probably due to steric
crowding causing similar conformational changes in the ring (Scheme 3).13
Thus, by employing these fluoro sugars as donors in BF3-etherate-pro-
moted couplings with Lipid A type disaccharide acceptor, a-(2-6)-linked
trisaccharides were obtained in high yields and with high a-selectivity.
Subsequent coupling to a Kdo 4,5-diol acceptor using the same donor(s)
afforded the a-(2-4)-linked tetrasaccharide. The high yields were accom-
plished by using a large excess (4 equivalents) of donor to compensate for
the high incidence of glycal formation. However, activation of the acceptors
as their triethylsilyl derivatives decreased the formation of glycal con-
siderably and consequently afforded increased yields of glycosides, although
the excess of donor were brought down to 2 equivalents.
Ichiyanagi et al. reported in 2011 on a new Kdo donor, but the only
novelty as compared to the derivatives used by Kusumoto and co-workers
discussed above was that benzoyl instead of benzyl protecting groups were
used in the 7,8-positions.14 Using (non-silylated) 4,5- or 7,8-diol Kdo
acceptors both a-Kdo-(2-4)-a-Kdo and the a-Kdo-(2-8)-a-Kdo dis-
accharides were produced using BF3-etherate as promoter, however, the
42 | Carbohydr. Chem., 2012, 38, 40–60

OTES
HO O OBn
O BnO
RCOO
TBSO
RCONH
BnO O O
RCOO TBSO CO2Bn
RCONH
OAll O
RCOOH = various fatty acids
HO O
BF3 . Et2O RCOO O
BnO OBn RCONH
89% BnO O
R 1O
RCOO
O
CO2Bn RCONH
R 1O
OAll
BnO OBn
R1 = TBS or F HO 4 steps
isopropylidene overall yield 73%
O
TESO CO2Bn
O
O
P O O
O RCOO O
BnO OBn RCONH
O
O BF3 . Et2O BnO
75% RCOO
O CO2Bn RCONH
O BnO OBn
OAll
HO
O O 1-O-phosphorylation
CO2Bn
and deprotection, 5 steps,
overall yield 15%
O
O
P O O Re LPS
O RCOO O
RCONH
BnO O
RCOO
RCONH
OAll
Scheme 3
stereoselectivity in the former coupling was only a/b 5:1 and the yield in the
latter was reported as ‘‘unsatisfactory’’ (Scheme 4).
Surprisingly, there was only one report before 2000 on the use of
thioglycosides as Kdo glycosyl donors, then with NIS/TfOH as promoter.15
In contrast, for sialylation reactions, thioglycosides have been used
extensively,16 e.g. O-xanthates promoted by DMTST17 or phenylsulfenyl
triflate18 and ethyl thioglycosides promoted by DMTST19 or IBr.20
Mannerstedt, Ekelöf and Oscarson reported on an investigation of
Kdo-thioglycosides as glycosyl donors testing DMTST, IBr/AgOTf and
NIS/AgOTf as promoters, however, the only acceptor investigated was
2-azidoethanol.21 The Kdo thioglycosides were found to be good donors
giving high yields of glycosides. The elimination reaction could be avoided
by performing the reactions at low temperature, which was possible through
the use of reactive promoters. Stereoselectivity was a problem, as would be
expected with a reactive spacer alcohol, a/b-mixtures were obtained often
Carbohydr. Chem., 2012, 38, 40–60 | 43

BzO OBz
BzO OBz
O
HO
O CO2Bn
O
CO2Me O BzO OBz
HO
HO
BF3 . Et2O
OAll O O
CO2Bn
BzO OBz
72% (α/β 5:1) OAll
O
O
O COOMe OBz
BzO
O
F
HO OH O CO2Bn
O O
O BF3 . Et2O O
O CO2Me HO
O
OAll O
O CO2Me
OAll
Scheme 4
with the non-natural b-anomer dominating. As discussed above also here a

4,5-O-isopropylidene group in the Kdo thioglycoside donor enhanced a-
selectivity.
Another type of Kdo donor was investigated by Reiner and Schmidt,22 a
phosphite donor, a type of donor which they had earlier developed and used
in formation of fructose glycosides also containing a quaternary aromatic
center (Scheme 5).
A 3-hydroxy analogue of Kdo was formed through a reaction between
2,3;5,6-di-O-isopropylidene-D-mannofuranose and the anion of methyl
glyoxylate diethyl mercaptal. Treatment of the obtained open dithioacetal
with NIS yielded the Kdo-analogue through ring formation followed by
hydrolysis of the intermediate thioglycoside. If the reaction was performed
at low temperature the ethyl thioglycoside could be isolated. The equatorial
3-OH group could be reduced to give the parent Kdo derivative, but was
also epimerized to yield a compound with an axial 3-OH group, which was
benzoylated, thus introducing a participating group that would ensure
a-selectivity in subsequent couplings. According to the authors the
thioglycoside was not a good donor, but was hydrolysed and transformed
into the anomeric diethyl phosphite derivate. This was used as donor pro-
moted by TMSOTf in couplings with two different 6-OH glucose acceptors
to give disaccharides in good yields (56 and 62%), and as expected with
complete a-selectivity. The auxiliary 3-hydroxyl group could then be
removed, in a three-step sequence with an overall yield of around 60%, to
give the corresponding protected Kdo-derivatives. Although a-selective this
synthesis suffers from the many transformations performed, giving a rather
low overall yield from the cyclized intermediate to the final (protected) Kdo
disaccharide.
44 | Carbohydr. Chem., 2012, 38, 40–60

O O
O O
O
NIS O
OH CO2Me MeCN O
O CO2Me
HO SEt O
76%
HO
SEt SEt
OH
4 steps
BnO O overall yield
O
O BnO 54%
O R BnO OMe O
O
O
CO2Me O OBz
O
O
CO2Me
O O
TMSOTf
BnO O 62%
BnO OP(OEt)2
BnO OMe
R = OBz 3 steps, overall

yield 64%
R=H
Scheme 5
Another way to Kdo oligosaccharides was described by Mlynarski and

Banaszek.23,24 Starting from D-mannose a 2-deoxy-heptono-1,5-lactone was
produced in a seven-step synthesis. The last carbon was introduced by
reacting the lactone with a dithiane to form a ketene dithioacetal in a
Horner-Emmons reaction. TMSOTf-promoted addition of alcohols to the
double bond then afforded a-glycosides both regio- and stereoselectively
(Scheme 6). High yields were obtained both with primary heptose- and Kdo
acceptors. NBS-promoted hydrolysis of the dithiane followed by treatment
with diazomethane then gave the target protected Kdo disaccharides in
about 50% yield.
Trichloroacetimidate donors of Kdo seem to be not useful in
oligosaccharide synthesis, but recently Fukase and co-workers reported
the use of N-phenyltrifluoroacetimidate Kdo donors in the formation
of a-Kdo-(2-6)-b-D-GlcNTroc disaccharides and a-Kdo-(2-6)-b-D-
GlcNTroc-(1-6)-a-D-GlcNAlloc trisaccharides (Scheme 7) for the synthesis
of Helicobacter pylori LPS structures.25,26 The structures synthesized are
similar to the ones shown in Scheme 3 (Reference 13) and the donor used is
the same except for the anomeric leaving group. After optimization they
found that triflic acid as promoter in acetonitrile gave high yields (62–88%)
and high stereoselectivity (68:32–100:0 a/b-ratio). Also conditions for per-
forming the glycosylations under microfluidic conditions were developed,
in which the a-selectivity of the reaction could be even further improved.
Several methods to form glycosides from easily obtained Kdo glycals by
addition to the double bond have been developed.1 This approach has the
advantage that the addition is almost exclusively transdiaxial affording the
desired a-glycoside selectively. Several electrophiles have been utilized to
Carbohydr. Chem., 2012, 38, 40–60 | 45

OBn
BnO
BnO OBn
BnO
BnO
O S
NBS, MeOH O
BnO
91% COOMe
BnO
S
OMe
ROH
TMSOTf OBn
OH BnO
BnO
BnO
BnO 72% S
O
O
COOMe BnO
BnO S
O
OMe BnO
BnO
1) NBS
BnO OBn O
2) CH2N2 COOMe
BnO BnO
50%
O
COOMe OMe
BnO
O
BnO
BnO
O
BnO COOMe
OMe
Scheme 6
activate the double bond, sulfenyl and selenyl reagents being the most
effective. So far, however, the inreactivity of the conjugated double bond
has required the use of quite reactive acceptors, e.g. primary spacer alco-
hols, for which the glycosylation yields are excellent. The stability of the
glycal is also showed by the fact that it is not activated by NIS alone.27
Though, if acetic or triflic acid is added in excess as an activator of NIS, the
addition can take place to produce the 3-iodo-glycosides in acceptable
yield.21,28
Tanaka, Takahashi and Takahashi developed this methodology further
to allow efficient synthesis of Kdo oligosaccharides (Schemes 8 and 9).29 The
glycals were formed from open dithianes, which were transformed into
sulphoxide/sulfide derivatives. Treatment of these with oxalyl chloride and
silver triflate first activated the sulphoxide to afford a cyclisation and then
subsequently the formed thioglycoside to yield the eliminated product. The
formed glycal derivatives were all benzyl protected thus increasing the
reactivity towards electrophiles as compared to earlier used acetylated
compounds. This together with the triflic acid activation proved to be an
effective combination for addition of even less reactive acceptors to the
Kdo-glycal. Employing a variety of monosaccharide acceptors the corre-
sponding 3-iodo-Kdo disaccharides were obtained in high yields and with
excellent but not complete a-stereoselectivity. The primary 8-hydroxyl
group of Kdo is known to be of surprisingly low reactivity. By using an
46 | Carbohydr. Chem., 2012, 38, 40–60

OH
BnO OBn
BnO OBn BnO O
O
O BnO
O
O TrocNH
O CO2Bn
CO2Bn OAll
O
TfOH, CH3CN O
O 86%, α/β 85:15
NPh BnO O
TfOH BnO
CH3CN CF3 OH TrocNH
70% OAll
α/β 95:5 BnO O
MPMO O
TrocNH
BnO O
BnO OBn ZO
O AllocNH
O OAll
O CO2Bn
O 8–9 steps
6–50% overall yield HpKdoLA
O and
BnO
HpKdoLAEA
MPMO O
TrocNH
BnO O
ZO
AllocNH
OAll
Scheme 7
BnO OBn
OBn
BnO BnO
O BnO
I
BnO COOEt O
BnO COOEt
NIS (1.5 equiv)
TfOH (1 equiv)
OH BnO O
BnO
BnO 73% BnO
O O
COOEt BnO COOEt
BnO
(1.5 equiv) OMe OMe

α/β 92:8
Scheme 8
open form 8-OH Kdo-precursor acceptor, the reactivity was enhanced to

allow an effective addition to the Kdo-glycal to produce an intermediate,
which could be transformed into a Kdo-Kdo disaccharide glycal as
discussed above. Iteration of this methodology made the synthesis of an
a-Kdo-(2-8)-a-Kdo-(2-8)-a-Kdo-(2-6)-b-D-Glc tetrasaccharide possi-
ble in high overall yield and stereoselectivity.
Carbohydr. Chem., 2012, 38, 40–60 | 47

OBn OBn
BnO BnO
BnO
BnO
O I
O
BnO COOEt COOEt
BnO
(COCl)2
NIS, TfOH O
BnO OH Bn O AgOTf
BnO 89% BnO
77%
OH COOEt OH COOEt
BnO BnO
S S O S S O
OBn OBn
BnO BnO
BnO BnO
I I
O O
COOEt BnO COOEt
BnO 1) NIS,
O α/β 95:5 TfOH O
BnO BnO
BnO 2) (COCl)2,
AgOTf BnO
O I
58% O
BnO COOEt BnO COOEt
OH O
BnO α/β 95:5
BnO O BnO
OC14H29
BnO
O
OBn
55% BnO COOEt
NIS, TfOH
OBn OH
BnO HO
BnO I HO
I
O O
COOEt
BnO HO COOH
1) H2, Pd(OH)
O 2) Ac2O, Pyr O
BnO HO
3) LiOH2
BnO I HO
I
O 47% O
BnO COOEt HO COOH
O O
BnO HO
BnO I HO
I
O O
BnO COOEt HO COOH
O α/β 86:14 O
BnO O HO O
OC14H29 OC14H29
BnO HO
OBn OH
Scheme 9
3 Glycero-D-manno-heptose-containing structures
The methodologies to produce glycero-D-manno-heptose, found in most
inner core LPS structures, most often as the L-glycero stereoisomer, were
mainly developed during the period 1980–2000 and were summarized in
the earlier review.1 During the last years a methodology described by Dasser
et al. in 1991 has become popular again.30 The methodology involves a
Grignard reaction on the 6-aldehyde function of a mannodialdo derivative
to give the vinyl alcohol which is converted to the corresponding heptose
48 | Carbohydr. Chem., 2012, 38, 40–60

derivative by dihydroxylation of the double bond, followed by periodate
cleavage and reduction. The methodology has a number of advantages,
notably complete L-selectivity irrespective of the protecting group pattern in
the mannose derivative, the use of a stable and commercial Grignard
reagent, which in most cases outweighs the disadvantage of the additional
steps required, and the synthesis can be performed on a large scale.31
Another reinvestigated methodology is the dihydroxylation of 6,7-ene
compounds, which as expected from Kishi’s empirical rule yields mainly
the D-isomer, while the L-isomer is available through a Mitsonobu
epimerization. However, the Wittig type ene-formation from the 6-aldehyde
precursor is contaminated with formation of 4,5;6,7-dienes and, thus, the
methodology competes well mainly when the D-isomer is wanted.32 Recently
a de novo synthesis of L-glycero-D-manno-heptose was published starting
from 1,3-O-isopropylidene-1,3-dihydroxyacetone and 2,2-dimethox-
yacetaldehyde. In an 11-step synthesis involving two stereoselective aldol-
condensations a partially protected heptose derivative was produced in
about 20% overall yield.33
In contrast to Kdo, for L-glycero-D-manno-heptose there is no
added complexity involved regarding their use in oligosaccharide synthesis.
Heptose donors behave more or less like hexose donors in glycosylation
reactions and the usual type of donors (thioglycosides, halides, and
acetimidates) have all been used with good success. Recent synthesis of L-
glycero-D-manno-heptose-containing oligosaccharides have to quite a large
extent concerned structures from the LPS inner core of Haemophilus influ-
enzae and Neisseria meningitidis, with a common branched tetrasaccharide
containing two L-glycero-D-manno-heptose residues, one Kdo moiety, and a
branching (1-4)-b-D-Glc residue linked to the first heptose (Figs. 1 and 2).
The 3,4-branched heptose structure is not trivial to synthesise, probably
due to steric crowding between the two equatorial substituents. Bernlind
and Oscarson circumvented this problem by using a 1,6-anhydro-heptose
derivative as acceptor.34 With the two substituents in diaxial positions the
glycosylations to form the branched structure was straight-forward. The
synthesis of the anhydro derivative has been improved and is now per-
formed from commercially available methyl a-D-mannopyranoside using
β-D-Glc
PEtN 1
| ↓
6 4
L-α-D-Hepp-(1→2)-L-α-D-Hepp-(1→3)-L-α-D-Hepp-(1→5)-α-Kdop
Fig. 1 Structure of the conserved inner core of H. influenzae LPS.
β-D-Glc α-Kdop
PEtN 1 2
| ↓ ↓
3 or 6 4 4
α-D-GlcNAcp-(1→2)-L-α-D-Hepp-(1→3)-L-α-D-Hepp-(1→5)-α-Kdop-(2→6)-LipidA
Fig. 2 Structure of the conserved inner core of N. meningitidis LPS.
Carbohydr. Chem., 2012, 38, 40–60 | 49

BDA-acetal protection and the vinyl carbon elongation approach discussed
followed by formation of the anhydro bridge by heating the methyl glyco-
side with FeCl3 in THF to produce the anhydro derivative in an overall
yield of B50% (Scheme 10), which then can be processed as earlier
described to 3,4-branched structures.
Orthogonally protected heptosyl thioglycoside donor have been synthe-
sised and introduced into the trisaccharide structure affording versatile
intermediates for continued synthesis of various N. meningitidis and
H. influenzae structures allowing almost any substituent pattern on the
second heptose and also the possibility for elongation at the reducing
end.35,36 These trisaccharide intermediates has been used in the synthesis
of spacer trisaccharides with variant phosphorylation pattern, a phos-
phoetanolamine in the 3 0 - or 6 0 -position, corresponding to Neisseria and
Haemophilus structures (Figs. 1 and 2 and Scheme 11).36
OH 1) a. (COCl)2, DMSO
OBn b. Et3N HO OBn
BnO O 2) CH2CHMgBr O
BnO
BnO 82% BnO
OMe OMe
OH
1) OsO4
FeCl3 NaIO4
O O
CH3CN H 2) NaBH4 H
reflux OBn OBn
O 88% O
OBn OBn
71%
OBn OBn
Scheme 10
AcO O BnO
OH
AllO OBz
OBz O
OBn O
BnO
BzO O ClAcO
O
BzO
OBz AcO O
Me2S2, Tf2O O
BnO CH2Cl2, 70% OBz O
OBn
AllO OBz
BzO O
BnO O O
BzO
ClAcO
OBz
SEt 8 steps
15 % overall yield
OH HO
O HO OH
HO
O O
HO
OH OH O
O P O O
NH2
HO O
O
HO OH
BocHN
HO
Scheme 11
50 | Carbohydr. Chem., 2012, 38, 40–60

A third heptose moiety was introduced into the 2 0 -position to give, after
opening of the anhydro bridge and introduction of an ethyl thioglycoside, a
tetrasaccharide building block donor corresponding to the outer part of the
conserved Haemophilus core structure. This tetrasaccharide donor has been
transformed into its 6-PEtN spacer derivative and also coupled to a Kdo
acceptor (Scheme 12).37,38
The corresponding branched trisaccharide (b-D-Glcp-(1-4)-[L-a-D-
Hepp-(1-3)]-L-a-D-Hepp) has also been synthesised by two Japanese
groups as a precursor for continued synthesis of LPS structures. Both these
groups show that it is possible to use a ‘‘normal’’ acceptor with a
4
C1-conformation in the production of the 3,4-branched structure, but care
has to be taken to find the right glycosylation conditions.
BzO
BnO
BzO BzO OBz
ClAcO
MeO OH BzO OBz BzO O
O O O BzO
O BzO
BzO BnO
MeO AcO ClAcO
O SEt MeO O
O O
O
O OBn O
OBz NIS, AgOTf
86% MeO AcO
BzO O O
O O
BzO
OBz Sc(OTf)2 OBz O OBn
Ac2O
BzO O
O
BzO
AcO OBz
OBz
AcO OBn AcO
O OBz
BzO O O AcO OBn
BzO O BzO O O
OBz NHCbz O
HO BzO O
AcO R OBz
AcO O NIS, AgOTf AcO O
R = OAc (97%) 72% AcO
ClAcO O O
AcO R = SMe (78%) O
ClAcO
BzO AcO NH
BzO O BzO
O O
BzO OBz O BzO O OBn
BzO HO BzO OBz
O
BnO COOMe BzO several steps
NIS, TfOH
60% O O HO
OH
HO OH
N OBn O
AcO H HO O O
OBz HO O
AcO OBn
O OH
BzO O O
BzO O O HO O
OBz O HO O
O O
AcO BocNHCH2CH2O O O
AcO O P H2 N
O HO HO
BnO COOMe
ClAcO O HO
AcO O O HO O
BzO HO OH
BzO O N OBn
H HO
BzO OBz
BzO
Scheme 12
Carbohydr. Chem., 2012, 38, 40–60 | 51

BnO BnO
OBz AgOTf OBz
BnO O BnO O
BzO O collidine BzO O
+ HO O O O
BzO 89% BzO
BzO O O
Br OBz
OSE OSE
1) NaOMe 4) Bu2SnO
BnO 2) Ac2O, pyridine 5) pMBnCl, TBAB
3) AcOH (80% aq) 6) BzCl, pyridine
BnO
OBn 7) CAN/CH3CN
OAc
AcO O O 61%
NIS, TMSOTf
AcO O O 54% BnO
OAc
AcO OMe BnO OBz
AcO O
RO BnO O O
BnO AcO
O BnO OBn OAc HO
BnO OBn O
BnO OSE
BnO LevO
R = Lev BzO OBz
NH2NH2-AcOH SMe
74% R=H AcO O
AcHN O O CCl3
O OBz
AcO NH
AcO
TMSOTf COOMe
AcHN O
48% BnO
AcO O
O BnO
C OBn
OAc OAc
OAc O AcO O
BzO O
AcO O O
BzO
AcO O AcO OMe
AcHN O
O
O BnO O
AcO OBz
AcO BnO OBn
COOMe
AcHN O BnO
AcO O
O
C
OAc O
OAc
Scheme 13
Thus, Hori et al.,39 when synthesising a structure from Campylobacter

jejuni LPS, also experienced that it is not possible to introduce the heptose
residue in the 3-position when there is a benzoyl-protected glucose moiety in
the 4-position. However, if the benzoyl protecting groups were changed to
acetates, then the glycosylation proceeded in an acceptable yield (54%,
Scheme 13).
The approach to the 3,4-branched structure used by Yamasaki and co-
workers was to use a mannose acceptor and first form the 3,4-branched
mannose structure, which was uncomplicated, and then subsequently make
the carbon-elongation at the trisaccharide level to produce the branched
heptoside (Scheme 14).40 However, with this late carbon elongation there
was a penalty to pay considering massive protecting group manipulation
and low-yielding reactions late in the synthesis.
The Yamasaki group has also synthesised 2,3-branched heptose struc-
tures corresponding to N. meningitidis LPS with an a-D-GlcNAc residue in
the 2-position (Scheme 15).41 The major problem here was not the
52 | Carbohydr. Chem., 2012, 38, 40–60

AcO
AcO OAc
AcO O
AcO 70% Ph O OH
O
O CCl3 BF3 . OEt2 O
O
NH AcO OMe
O OH AcO O
Ph
O O AcO OAc
HO AcO
TBDPSO
OMe 1) Ac2O, pyridine OAc
2) AcOH (70% aq) HO O
3) TBDPSCl, O
imidazole OMe
AcO
72% AcO O
NH AcO OAc
AcO OAc OAc AcO
O AcO O CCl3 BF3 . OEt2
AcO O O 59%
OAc
OAc
TBDPSO
AcO OAc OAc OAc
O AcO O O
AcO O O O
OAc AcO OMe
AcO
1) MeOH, Et3N, H2O AcO O
2) BnBr, NaH, DMF AcO OAc
3) TBAF, THF 61%
AcO
HO
BnO OBn OBn OBn
O BnO O O
BnO O O O
OBn BnO OMe
BnO
BnO 1) (COCl)2, 4) Ac2O, pyridine
O
DMSO 5) OsO4, NaIO4
BnO OBn
51% 2) Et3N 6) NaBH4
BnO
3) CH2CHMgBr
HO
HO
BnO OBn OBn OBn
O BnO O
O
BnO O O O
OBn BnO OMe
BnO
BnO O
BnO OBn
BnO
Scheme 14
formation of the branched structure but rather the a-stereoselective intro-

duction of the GlcNAc moiety. A 6-O-acetyl-2-azido-3,4-di-O-benzyl-2-
deoxy-b-D-glucopyranosyl trichloroacetimidate donor was utilized, but
after optimisation of reaction conditions the best yield of a-anomer
obtained was 58% (a/b 2.6/1). After protecting group manipulation a 3-OH
acceptor was produced which was glycosylated with a lactosyl tri-
chloroacetimidate donor to yield the target 2,3-branched structure. In this
coupling both the yield of the desired a-anomer (77%) and the a/b-ratio
(9.6/1) was much better.
Carbohydr. Chem., 2012, 38, 40–60 | 53

OAc
BnO O OAc
1) TFA (90% aq)
BnO O CCl3 O
BnO 2) TESCl, pyridine
N3 BnO
NH TMSOTf 3) Ac2O. pyridine
AcO N3
Et2O 4) TFA (90% aq)
O
AcO MeO AcO
80% O 84%
AcO O
MeO OH α/β 2.6:1 O
O O
O MeO
OMe
MeO OMe
NH
BnO OBn OBn
O BnO O CCl3
O
BnO O
OBn
OAc OBn OAc
BnO O BnO O
BnO BnO
85%
AcO N3 AcO N3
α/β 9.6:1
AcO O AcO O
O TMSOTf (0.04 eq), Et2O AcO O
AcO
HO O
BnO OBn BnO
OMe O BnO OMe
BnO O O
OBn
OBn
Scheme 15
The Oscarson group has also synthesized Neisseria structures starting

from the 1,6-anhydro heptose intermediates discussed above.42 A GlcNAc
residue was introduced into the 2 0 -position and the formed tetrasaccharide
was transformed, as for the corresponding Haemophilus structure, into a
tetrasaccharide thioglycoside donor and further into a 6-PEtN spacer
derivative (Scheme 16). As opposed to the findings of Yamasaki (Scheme
15), no stereoselectivity problems were encountered in model couplings
between a monosaccharide thioglycoside heptose acceptor and an azide
trichloroacetimidate donor (see also below, Scheme 17). However, when the
obtained disaccharide was tried as a donor in a coupling to the earlier used
disaccharide acceptor (Scheme 11) this failed. Hence, the planned [2 þ 2]
strategy had to be abandoned. Instead a [1 þ 3] strategy was tried using the
same trisaccharide acceptor as was earlier used for Haemophilus structures
(Scheme 12). Surprisingly only NIS/AgOTf-promoted glycosylations
with the thioglycoside azide donor worked. The yield was excellent but the
stereoselectivity low.
Recently, the synthesis of the linear Neisseria tetrasaccharide structure
a-D-GlcNAc-(1-2)-L-a-D-Hep-(1-3)-L-a-D-Hep-(1-5)-a-Kdo as its
5-aminopentyl glycoside was reported by Seeberger and co-workers.43 In
this case, perhaps due to the fact that the 3,4-branched structure was not
targeted, the tetrasaccharide could be obtained through a [2 þ 2] block
synthesis using a a-D-GlcN3-(1-2)-LD-Hep disaccharide donor in the
last glycosylation (Scheme 17). In the synthesis trichloro- and N-phenyl
54 | Carbohydr. Chem., 2012, 38, 40–60

BnO
MeO CAO OH
O O OAc
O N3 OAc
O
MeO AcO O BnO OAc
O MeO CAO O
OBz O O 5 steps
OBn O
BzO O O
O 47%
BzO NIS, AgOTf
OBz MeO AcO O
95% O
OBz O
OAc α/β 1.5:1 OBn
BzO O
AcO O O
AcO SEt BzO
OBz
N3
OH HO
OBz AcO O HO OH
AcO OBn HO O
BzO O HO O
BzO O O OH O
5 steps
OBz O
33% HO NH2 O
AcO SEt HO O
AcO O O O
O
CAO O HO OH
HO P O
AcO OAc AcHN OH
O
N3 OAc OCH2CH2NHBoc OH
OAc
Scheme 16
trifluoroacetimidate donors promoted by TMSOTf were employed, since

the corresponding thiophenyl donors promoted by NIS/TMSOTf failed to
give the desired glycosides. It is interesting to notice that Bernlind and
Oscarson, in coupling of ethyl thioglycoside heptose derivatives to a very
similar Kdo acceptor, found that this coupling worked well with a 2-O-
benzyl protecting group in the heptose donor, but failed with a 2-O-benzoyl
protected donor.34
As mentioned, D-glycero-D-manno-heptose is more rarely found in the
core structure of LPS, but some examples reported are LPS structures from
Haemophilus, Klebsiella and Plesimonas, in the latter case as the even more
rarely found b-glycoside. Bernlind and Oscarson synthesized a D-glycero-
a-D-manno-heptose-containing pentasaccharide structure from H. ducreyi as
discussed in the earlier review,44 and Gurjar and Talukdar have published a
synthesis of an a-DD-Hep-(1-2)-a-DD-Hep disaccharide motif found in
K. pneumonia (Scheme 18).45 In the latter publication the D-glycero-D-manno-
heptose derivatives were obtained from a 6,7-olefin precursor through a
Sharpless asymmetric dihydroxylation reaction affording a 9:1 D/L-mixture,
from which the D-isomer could be isolated as its 6,7-isopropylidene deri-
vative in 47%. In this synthesis, a bit surprisingly, a 1-O-acetate donor was
used, activated by BF3-etherate, why the yield in the coupling reaction was
only 26%.
Crich and Banerjee in their synthesis of Plesimonas shigelloides LPS-
structures also utilized a 6,7-olefin precursor, but the dihydroxylation
was performed using OsO4 (Scheme 19).32,46 The anomeric thiophenyl
group was not affected under these oxidating conditions. As expected from
Carbohydr. Chem., 2012, 38, 40–60 | 55

O
O
BzO HO
BnO OH O
BnO CO2Me
PBBO O
BzO AcO + O NBnCBz
+
SAr BnO OAc 5
OAc
O PBBO O
AcO
AcO LevO
N3 O CCl3
CCl3 TMSOTf,
O
TMSOTf, CH2Cl2, 88% NH
Et2O, 68% NH
AcO
OAc BnO OAc
AcO O PBBO O
AcO R2O O O
N3
BzO R2 = Lev O
BnO O 73% O
R2 = OH BnO
O CO2Me
PBBO
BzO O NBnCBz
5
R1 TMSOTf,
R2 = SAr
66% Et2O/CH2Cl2
R2 = OC(NPh)CF3 AcO
OAc 72%
BnO
PBBO O
O O O
BzO O
PBBO O
O BnO CO2Me
BnO O
BzO OAc O NBnCBz
O
N3 OAc
OAc
Scheme 17
BnO HO
BnO
BnO HO
BnO OAc OH
OAc
O HO O
BnO O BnO
BnO HO
BnO
1) NaOMe
BF3 . Et2O BnO HO
BnO AcO 2) H2, Pd/C O
CH2Cl2 BnO O HO
BnO 52%
OH 26% O HO O
O BnO
BnO BnO HO
BnO
OMe OMe
OMe
Scheme 18
Kishi’s empirical rule this afforded an excess of the D-isomer (D/L 5:1 at
0 1C). According to the authors asymmetric dihydroxylation did not
improve these results. To get good b-selectivity in following coupling
reactions with these DD-Hep derivatives, the authors utilized their
56 | Carbohydr. Chem., 2012, 38, 40–60

OBn
HO OBn OBn
PMBnO O
O + O
PMBnO BnO
BnO
BnO
1) DMSO, (COCl)2, 63% SPh 14%
SPh SPh
Et3N
2) CH2PPh3Br, BuLi OsO4 0 °C
NMNO
HO HO
TBDPSO
HO OBn HO OBn
HO OBn
TBDPSCl PMBnO O PMBnO O
PMBnO O
BnO + BnO
BnO
66% SPh 13% SPh
SPh
DDQ 84%
TBDPSO
TBDPSO O OBn
PMPh OBn
O
O OBn 1) BSP, O
O O
PMPh O TTBP, Tf2O BnO
O BnO
BnO BnO
OMe
OBn 81%
SPh
HO O
2) BnO
BnO
OMe
Scheme 19
own methodology originally developed for the stereoselective formation of

b-D-mannopyranosides, i.e. pre-formation of an a-triflate, stabilized by a
4,6-acetal group, and a subsequent SN2-reaction to displace the anomeric
triflate with an acceptor. DDQ-oxidation of the 4-O-p-methoxybenzyl
group in the heptose formed the necessary 4,6-O-benzylidene acetal through
participation of the neighbouring free 6-OH group.
Glycosylation with these donors using 1-benzenesulfinyl piperidine/triflic
anhydride as promoter at 60 1C gave, as earlier found with mannose
donors, high yields and excellent b-selectivity. The Plesimonas structure
targeted does not only contain a D-glycero-D-manno-heptose residue but
also a 6-deoxy-D-manno-heptose residue, both of them b-linked. To
accomplish the latter structural motif a heptose thioglycoside containing a
4,6-O-[1-cyano-2-(2-iodophenyl)]ethylidene acetal was synthesized. When
used as a donor this 4,6-acetal ensures b-selectivity as discussed and can
then subsequently be fragmented by treatment with Bu3SnH and AIBN to
afford a 6-deoxy derivative (Scheme 20). By combined use of these two
b-directing D-heptose thioglycoside donors an efficient synthesis of a linear
tetrasaccharide from the target LPS structure, a-L-Rha-(1-3)-D-b-D-Hep-
(1-3)-6d-b-D-Hep-(1-4)-a-L-Rha, could be accomplished.
In a following paper Crich and Li used the same approach to synthesize
di- and trimers of the disaccharide repeating unit of a glycan from a
surface-layer glycoprotein of Bacillus thermophilus, [-4-a-L-Rha-(1-3)-D-
b-D-Hep-(1-]n.47 The homologation procedure to heptose derivatives
was improved by using a 3,4-BDA-acetal in the mannose precursor,
Carbohydr. Chem., 2012, 38, 40–60 | 57

I
I CN BnO
CN BnO
O OBn OMe
O OBn 1) DPSO, O
O O
O O TTBP, Tf2O
NapCH2O O
NapCH2O O
86% O
SPh 2) OMe
1) Bu3SnH, AIBN
O
HO 2) DDQ
O O CN BnO
O MeO
OBn
O O O
HO O
BnO O
57% O
O OBn
Ph O 1) BSP, TTBP, Tf2O
O 2)
NapCH2O
3) DDQ
SPh BnO
CN
O OMe
OBn
SPh O
O O
O BnO O O
BnO O OBn
Ph O O
O O O
O
HO 88%
1) BSP, TTBP, Tf2O

2) CN BnO
O
OBn OMe
O O O
BnO O O
O OBn O
Ph O O
O 73%
O
1) NaOMe
O 2) TFA
BnO
3) H2, Pd(OH)2/C
O O HO
HO OH OMe
HO OH HO O O
HO O O O
O HO OH
O 81%
HO
HO OH
Scheme 20
which prevented the formation of the elimination side product in the

Swern oxidation/Wittig reaction sequence and thus improved the yield
with about 20% over the two steps. However, the dihydroxylation step
with the 3,4-BDA-acetal derivative was found to be much less selective
for the D-glycero isomer. Following protecting group manipulation on
the DD-derivative afforded a suitably protected thioglycoside heptosyl
donor with the required 4,6-O-benzylidene acetal. Coupling of this with
either a methyl rhamnoside or a phenyl thio-rhamnoside acceptor gave a
starting disaccharide and an elongating disaccharide building block,
respectively, both with complete b-selectivity. These building blocks were
then combined through NIS/AgOTf promoted couplings to give the target
tetra- and hexasaccharides in high yield and with complete a-selectivity
(Scheme 21).
58 | Carbohydr. Chem., 2012, 38, 40–60

BnO
O OBn
Ph O
O
1) BSP, Tf2O, NapO
1-octene SPh 1) 4-NO2PhSCl,
2) SPh AgOTf 2) OMe
O
HO 73% O
O HO
O
O O
BnO Ph Ph
BnO
OBn SPh Ph Ph
Ph O OBn OMe
O Ph O
O O O
NapO O O O
RO O
O O O O
R = Nap, 81%
Ph Ph DDQ R = H, 88% Ph Ph
NIS/AgOTf
BnO
OBn OMe
O
Ph O
BnO O O
OBn O O
Ph O
O O O
O O
RO O 1, NIS/AgOTf Ph Ph
O O 64%
R = Nap, 85%
DDQ Ph Ph
R = H, 74% BnO
OBn OMe
O
Ph O
BnO O O
OBn O O
Ph O
O O O
O O
RO O Ph Ph
O O
Ph Ph
2
Scheme 21
References
1 J. Hansson and S. Oscarson, Curr. Org. Chem., 2000, 4, 535.
2 P. Kosma, Curr. Org. Chem., 2008, 12, 1021.
3 P. Kosma, ‘‘Chemical synthesis of the core oligosaccharide of bacterial
lipopolysaccharide’’, in Microbial Glycobiology (Ed. A. P. Moran), 2009, 429.
4 L.-S. Li and Y.-L. Wu, Curr. Org. Chem., 2003, 7, 447.
5 A. Banaszek and J. Mlynarski, Studies in Natural Prod. Chem., 2005, 30, 419.
6 M. A. Ghalambor and E. C. Heath, Biochem. Biophys. Res. Commun., 1963, 10, 340.
7 J. W. Cornforth and M. E. Firth, Biochem. J., 1958, 68, 57.
8 P. Kosma, A. Reiter, A. Hofinger, L. Brade and H. Brade, J. Endotoxin Res.,
2000, 6, 57.
9 P. Kosma, A. Hofinger, A. Muller-Loennis and H. Brade, Carbohydr. Res.,
2010, 345, 704.
10 M. Blaukopf, B. Mueller, A. Hofinger and P. Kosma, Eur. J. Org. Chem., 2012, 119.
11 M. Imoto, N. Kusunose, Y. Matsuura, S. Kusumoto and T. Shiba, Tetrahedron
Lett., 1988, 29, 6277.
Carbohydr. Chem., 2012, 38, 40–60 | 59

12 M. Imoto, N. Kusunose, S. Kusumoto and T. Shiba, Tetrahedron Lett., 1988,
29, 2227.
13 H. Yoshizaki, N. Fukuda, K. Sato, M. Oikawa, K. Fukase, Y. Suda and S.
Kusumoto, Angew. Chem. Int. Ed., 2001, 40, 1475.
14 T. Isciyanagi, M. Fukunaga, R. Tagashira, S. Hayashi, M. Nanjo and R.
Yamasaki, Tetrahedron, 2011, 67, 5964.
15 G. J. P. H Boons, F. L. Delft, P. A. M van Klein, G. A. van der Marel and J. H.
van Boom, Tetrahedron, 1992, 48, 885.
16 R. H. Halcomb and M. D. Chappell, J. Carbohydr. Chem., 2002, 21, 723.
17 A. Marra and P. Sinay, Carbohydr. Res., 1990, 195, 303.
18 V. Martichonok and G. M. Whitesides, J. Org. Chem., 1996, 61, 1702.
19 A. Hasegawa, H. Okhi, T. Nagahama, H. Ishida and M. Kiso, Carbohydr. Res.,
1991, 212, 277.
20 A. Meijer and U. Ellervik, J. Org. Chem., 2002, 67, 7407–7412.
21 K. Mannerstedt, K. Ekelöf and S. Oscarson, Carbohydr. Res., 2007, 342,
631–637.
22 M. Reiner and R. R. Schmidt, Tetrahedron Asymm., 2000, 11, 319–335.
23 J. Mlynarski and A. Banszek, Org. Lett., 1999, 1, 1709–1711.
24 J. Mlynarski and A. Banszek, Tetrahedron Asymm., 2000, 11, 3737–3746.
25 A. Shimoyama, Y. Fujimoto and K. Fukase, Synlett., 2011, 16, 2359.
26 A. Shimoyama, A. Saeki, N. Tanimura, H. Tsutsui, K. Miyake, Y. Suda, Y.
Fujimoto and K. Fukase, Chem. Eur. J., 2011, 17, 14464.
27 M. Brenken, Doctoral Thesis, 1991, Hamburg University
28 P. Kosma, H. Sekljic and G. Balint, J. Carbohydr. Chem., 1996, 15, 701.
29 H. Tanaka D. Takahashi, and T. Takahashi, Angew. Chemie Int. Ed., 2006, 45, 770.
30 M. Dasser, F. Chretien and Y. Chapleur, J. Chem. Soc. Perkin Trans., 1990, 1, 3091.
31 R. Yamasaki, A. Takajyo, H. Kubo, T. Matsui, K. Ishii and M. Yoshida,
J. Carbohydr. Chem., 2001, 20, 171.
32 D. Crich and A. Banarjee, Org. Lett., 2005, 7, 1395–1398.
33 T. Ohara, A. Adibekian, D. Esposito, P. Stallforth and P. H. Seeberger, Chem.
Commun., 2010, 46, 4106.
34 C. Bernlind and S. Oscarson, J. Org. Chem., 1998, 63, 7780.
35 E. Segerstedt, K. Mannerstedt, M. Johansson and S. Oscarson, J. Carbohydr.
Chem., 2004, 23, 443.
36 J. D. M. Olsson and S. Oscarson, Carbohydr. Res., 2010, 345, 1331.
37 K. Mannerstedt, E. Segerstedt, J. Olsson and S. Oscarson, Org. Biomol. Chem.,
2008, 6, 1087.
38 K. Mannerstedt, Doctoral Thesis, 2006, Stockholm University
39 K. Hori, N. Sawada, H. Ando, H. Ishida and M. Kiso, Eur. J. Org. Chem.,
2003, 3752.
40 K. Ishii, Y. Esumi, Y. Iwasaki and R. Yamasaki, Eur. J. Org. Chem., 2004, 1214.
41 H. Kubo, K. Ishii, H. Koshino, K. Toubetto, K. Naruchi, and R. Yamasaki,
Eur. J. Org. Chem., 2004, 1202
42 J. D. M. Olsson and S. Oscarson, Tetrahedron Asymm., 2009, 20, 879.
43 Y. Yang, C. E. Martin and P. Seeberger, Chem. Sci., 2012, 3, 896.
44 C. Bernlind, S. Bennett and S. Oscarson, Tetrahedron Asymm., 2000, 11, 481.
45 M. K. Gurjar and A. Talukdar, Tetrahedron, 2004, 60, 3267.
46 D. Crich and A. Banarjee, J. Am. Chem. Soc., 2006, 128, 8078.
47 D. Crich and M. Li, J. Org. Chem., 2008, 73, 7003.
60 | Carbohydr. Chem., 2012, 38, 40–60

Synthetic glycopeptides in vaccine
development and antibody epitope
mapping
Ulrika Westerlind
DOI: 10.1039/9781849734769-00061
Over a number of years, synthetic methods for the preparation of glyco-

peptide based vaccines have been refined to target the aberrant glycosylated
and extensively over-expressed membrane-bound mucin MUC1 glycopro-
tein. On the epithelial cell surface, MUC1 forms tumor specific epitopes
consisting of both the aberrantly short saccharides e.g. TN, T, sialyl-TN and
sialyl-T and the exposed MUC1 tandem repeat peptide backbone. By syn-
thesizing defined glycopeptide structures consisting of the MUC1 tandem
repeat region, antibodies selective to the tumor cell surface could be formed.
However the tumor-selective MUC1 glycopeptides are only moderately
immunogenic and additional stimulation is necessary to generate a humoral
immune response. Recent synthesis of MUC1 glycopeptides conjugated to
different immuno-stimulants and promising results of evaluated antibodies
induced with the synthetic MUC1 vaccines will here be highlighted.
1 Introduction
Specific immune recognition, where both the glycan and the peptide back-
bone contribute to the binding epitope is of particular interest for devel-
opment of safe immunotherapy and immunodiagnostics. Since the
discovery by Springer, that glycoproteins on the outer cell-membrane of
epithelial tumor cells have an altered glycosylation consisting of the
Thomsen-Friedenreich (T-) antigen and its precursor the TN-antigen
structure, the synthesis and evaluation of anti-tumor vaccines have been
dominating in glycopeptide based vaccine research.1 Recent advancement in
miniaturized bioanlytical techniques and improvements in glycopeptide
synthesis have further made it possible to efficiently analyze antibody
binding-epitopes in a microarray format.2
Mucins are a class of extensively glycosylated proteins expressed on the
surface of epithelial cells or secreted to function in mucus. Among them the
membrane-bound mucin 1 (MUC1) is expressed on almost all epithelial
tissues. Changes of protein glycosylation on the epithelial tumor cell-
surface, with concomitant down-regulation of glycosyl transferases, in par-
ticular the core 2 b-1,6-N-acetylglucosaminyltransferase, and up-regulation
of sialyltransferases together with increased MUC1 protein expression levels
result in the formation of tumor specific epitopes.3 The aberrant under-
glycosylation of the mucin extracellular domain further lead to the exposure
of its peptide backbone. The specific tumor epitopes formed therefore consist
Gesellschaft zur Förderung der Analytischen Wissenschaften e.V., ISAS – Leibniz Institute for
Analytical Sciences, Otto-Hahn-Str. 6b, D-44227 Dortmund, Germany.
E-mail: ulrika.westerlind@isas.de
Carbohydr. Chem., 2012, 38, 61–74 | 61

of both the aberrant short saccharides e.g. TN, T, sialyl-TN and sialyl-T and
the exposed MUC1 tandem repeat peptide backbone (PAHGVTSAPDTR-
PAPGSTAP) (Fig. 1). Moreover these tumor-associated glycan structures on
MUC1 play important roles in adhesion/anti-adhesion binding events pro-
moting tumor progression and tolerance by the immune system.4 By inducing
sufficiently strong immune reactions specific to MUC1 tumor-associated
antigens, it should be possible to break the immune tolerance against these
structures. A number of synthetic vaccines with the moderately immunogenic
MUC1 tumor associated glycopeptide epitopes as target have recently been
prepared and employed with different immuno-stimulants. Among the
immuno-stimulants, carrier proteins like Tetanus Toxoid protein (T.Tox.)
and bovine serum albumin (BSA) and the Toll-like receptor 2 (TLR2) ligand,
tripalmitoyl-(S)-glyceryl lipopeptide Pam3-Cys-Ser-(Lys)4 (Pam3CSK4), have
shown to be particular promising during application in MUC1 anti-tumor
vaccines.5 Multivalent glycopeptide dendrimer vaccines were recently pre-
pared and are currently employed in immunological studies.6
2 Protein conjugate vaccines

Synthetic MUC1-Tetanus Toxoid conjugate vaccines have proven to be
particularly interesting. The Tetanus Toxoid protein has already been applied
to humans in several vaccines towards bacterial and viral targets. Recent
immunizations of MUC1 vaccines in mice have resulted in highly specific and
unusually strong immune responses. Furthermore several antibodies induced
by these vaccines showed specific recognition of tumor cells.5b,c
The MUC1 tandem repeat glycopeptides containing sialyl-TN-, T- or
di-fluoro-T-antigen and a triethylene glycol spacer with a free amine in
N-terminus were synthesized on solid-phase according to the Fmoc strategy
using Fmoc amino acids and Fmoc glycosyl amino acid building blocks in a
stepwise fashion. Trityl resin was used in order to avoid diketopiperazine
formation. Employing TFA the fully assembled peptides were cleaved from
the resin with simultaneous removal of the peptide side chain protective
groups. The sialyl-TN-disaccharide was then deprotected by hydrogenation
using palladium on charcoal and subsequent deacetylation using NaOMe in
methanol resulting in the formation of the sialyl-TN-MUC1-glycopeptides
1, 4 and 5. The T- and di-fluoro-T-glycopeptides were directly treated with
base to give the deacetylated compounds 2 and 3. The deprotected MUC1-
glycopeptides reacted with diethyl squarate by stirring in a mixture of
ethanol and water at pH 8 to yield the activated monoamide constructs. The
squarate monoamide-glycopeptides were conjugated to the Tetanus Toxoid
protein in a sodium phosphate buffer at pH 9 and afforded vaccines 6–10
carrying on average more than 20 MUC1-glycopeptides conjugated to the
protein (Scheme 1). Immunological evaluation of MUC1 glycopeptide
vaccines containing a sialyl-TN, a T-antigen or a di-fluoro-T-antigen on
different positions on the tandem repeat (6–10) showed that extraordinary
high antibody titers were induced in almost all of the immunized mice.
Evaluation of the generated antibodies by ELISA and neutralization
experiments as well as by glycopeptide microarrays gave evidence that
specific immune responses were elicited towards the MUC1 antigens present
62 | Carbohydr. Chem., 2012, 38, 61–74

Fig. 1 Tumor-associated glycosylation on MUC1 tandem repeat peptides.
Carbohydr. Chem., 2012, 38, 61–74 | 63

A: Fmoc-cleaveage:
Piperidin (20 %) in NMP OH F HO F
Di-Fluoro-T = O O
N O SPPS B: Amino acid-coupling: HO O
Fmoc
Fmoc-AA-OH, HBTU, HOBt, DIPEA/DMF HO AcHN
O Ph Ph
C: Capping: HO OH CO2H
= TentaGel R Trityl HO OH OH HO OH
Cat. HOBt, Ac2O, DIPEA in NMP O
Loading: 0.18 mmol/g AcHN O T= O O
Resin-cleavage: HO HO HO O
Sialyl-T = O HO AcHN
TFA, TIS, H2O (15:0.9:0.9), 2 h N HO
1: Benzylester-cleavage: AcHN
H2, Pd/C (10%), 15 h in MeOH
O O
O NH2 R1 2: NaOMe, MeOH, 20 h R2
N NH
O O HO O
O H H O H O H O H O O H O H O
N N N N N N N N N N N R4
N N N N N N N N N
H H H H H H H H O O
O O O O O O O O O O
OH O O
HO
NH
Compound: 1–5 R3
H2N NH
64 | Carbohydr. Chem., 2012, 38, 61–74

O O
EtOH, H2O 1:1, 2h
EtO OEt
O O
O
T.Tox.* N N
H H T.Tox. NH2
O O R1 R2
O N NH
O HO O
O H H O H O H O H O O O H O
H
N N N N N N N N N N N R4
N N N N N N N N N
H H H H H H H H O O
O O O O O O O O O O
OH O O n
Compound: 6–10 HO
NH
R3
2 HN NH
1, 6: R1 = Sialyl-TN, R2 = H, R3 = H, R4 = OH
4 4, 9: R 1 = H, R2 = Sialyl-TN, R3 = H, R4 = P-A-CO2H
2, 7: R1 = T, R2 = H, R3 = H, R = OH
4
3, 8: R1 = Di-Fluoro-T,R2 = H, R3 = H R4 = OH 5, 10: R1 = H, R2 = H, R3 = Sialyl-TN, R = P-A-CO2H
Scheme 1 Synthesis of MUC1 Tetanus Toxoid protein conjugate vaccines.

in the vaccines. FACS analysis of serum antibodies induced by the MUC1
vaccines 6–10 showed all that they recognize MCF-7 breast cancer cells. The
serum from mice immunized with vaccine 9 was further evaluated through
mammary carcinoma tissue staining experiments. A gradual increase of the
tissue staining was found, resulting in a very strong binding to the advanced
G3 phase tumor tissue. Taken together these results show that a selective
immune response discriminating between healthy and diseased tissue can be
efficiently obtained.5a–c
In a different study MUC1 glycopeptides conjugated to a bovine serum
albumin (BSA) immune carrier protein were evaluated.5d,7 In accordance
with the above described T.Tox. vaccines, the MUC1 glycopeptides were
conjugated to the protein through coupling to diethylsquarate. The MUC1
peptides were glycosylated with the TN- and T- antigen on Thr9 in the
immuno-dominant PDTR region or on Ser15 in the GSTA region of the
tandem repeat. Immunological evaluation showed that the elicited antibody
titers were depending on the position glycosylated. Vaccines glycosylated
with TN or T (26, 29, 30) in the MUC1 PDTR region induced a strong
immune response. Glycosylation with the T-antigen or TN of Ser15 resulted
in medium or weak antibody titers (25, 27, 28). No glycosylation of the
MUC1 backbone (24) resulted in weak immune responses.7 Triggered by
these results, a number of sialyl-TN and 2,6-sialyl-T glycopeptide BSA
constructs (31–36) modified on Thr9 in the PDTR domain and on Ser15 in
the GSTA region were prepared. Immunization in mice resulted in strong
immune responses in all applied sialyl-TN and 2,6-sialyl-T MUC1 vaccines
(31–36).5d Studies of antibody isotypes from the antibody serum showed
that the IgG1 isotype is predominant in each case, but some IgM, IgA,
IgG2a,b and IgG3 antibodies are also present. FACS analysis to study
binding recognition to MCF-7 breast tumor cells showed that antibodies
induced by the non-glycosylated vaccine 24 showed hardly any binding
recognition to the tumor cells, vaccine 31 carrying a sialyl-TN on Ser-15
surprisingly only showed weak binding. Antibody sera induced by vaccines
27, 28, 30 and 32–36 showed a strong binding to the tumor cells. Sera from
vaccines 25, 26 and 29 were not tested in the FACS analysis binding studies.
From these studies it could be concluded that compared with Tetanus
Toxoid conjugates, the BSA vaccines are more dependent on glycosylation
and MUC1 backbone conformation in order to efficiently induce a strong
immune response. However, the BSA immune carrier is cheaper than the
Tetanus Toxoid protein carrier and as shown here, it can in many occasions
induce strong and tumor selective immune responses (Scheme 2).
3 Build-in adjuvant vaccines

A number of two- or three-component vaccines containing a Toll-like
receptor two (TLR2) ligand have been synthesized and evaluated.5e,f,g These
self-adjuvanting vaccines avoid invoking an immune response to the
immune-carrier, a problem commonly seen employing protein conjugate
vaccines. The synthesis of two-component vaccines with a Pam3Cys TLR2
ligand connected to TN, T and sialyl T MUC1 glycopeptides was recently
described (Scheme 3). The Pam3CSK4 lipopeptide fragment 40 was
Carbohydr. Chem., 2012, 38, 61–74 | 65

NH2
R2 R1
N NH
O 3 HO O O
O O O O O O O O
H H H H H H H
O NH N N N N N N N N N N N
N N N N N N N N OH
H H H H H H H
O O O O O O O O O O O O
OH O HO n
HO
NH
HO
H2N NH OH CO2H
Compound 11–23 HO
O
O O AcHN O
HO HO HO
HO OH CO2H
O
OH HO OH EtO OEt HO O
OH OH AcHN
HO AcHN O
T= O O STN = HO
TN = O HO O EtOH, H2O 1:1, 2h OH OH HO
HO
HO AcHN O O
AcHN HO O
O O
HO AcHN
66 | Carbohydr. Chem., 2012, 38, 61–74

BSA NH2 2,6-ST =
BSA * N NH
H
R2 R1
N NH
O 3 HO O O
O O O O O O H O H O
H H H H H
O NH N N N N N N N N N N N
N N N N N N N N OH
H H H H H H H
OH O HO n
HO
Compound 24–36 NH
H 2N NH
11, 24: R1 = H, R2 = H 15, 28: R1 = TN, R2 = H 18, 31: R1 = STN, R2 = H 21, 34: R1 = STN, R2 = TN
12, 25: R1 = T, R2 = H 16, 29: R1 = H, R2 = TN 19, 32: R1 = STN, R2 = T 22, 35: R1 = 2,6-ST, R2 = H
13, 26: R1 = H, R2 = T 17, 30: R1 = TN, R2 = TN 20, 33: R1 = 2,6-ST, R2 = TN 23, 36: R1 = 2,6-ST, R2 = T
14, 27: R1 = TN, R2 = T
Scheme 2 Synthesis of MUC1 BSA protein conjugate vaccines.

R
N NH
O O HO HO
O O O O O O O O
H H H H H H H
N N N N N N N N N N N OH
N N N N N N N N N
O H H H H H H H H
OH O HO
O HO
NH MUC1 epitope HO CO2H
OH
H2N NH HO
O O
AcHN
1. Pam3Cys lipopeptide 40, OH OH OH HO O
OH OH HO
NH2 HOAt, HATU, NMM in DMF, 15 min O
OH OH HO
HO O O
2. 37, 38 or 39 in DMF 20 h HO O O O
AcHN HO O
3. TFA/TIS/H2O (10:1:1), 1.5 h HO AcHN
HO AcHN
37 38 39
R
N NH
O O HO HO
O O O O O O O O
H H H H H H H
N N N N N N N N N N N OH
N N N N N N N N N
O H H H H H H H H
OH O HO
O HO
NH2 NH2 NH MUC1 epitope
O H2N NH
O HO
OH OH CO2H
O O O C15H31 O HO
H H
N N O
HN N N N S AcHN
H H H OH OH HO O
O O O NH O OH OH HO OH OH OH HO
O O
C15H31 C15H31 R= HO O O O O
O HO O HO O
H2N AcHN AcHN
NH2 HO AcHN HO
Toll-Like receptor Ligand Pam3CSK4 41 42 43
Carbohydr. Chem., 2012, 38, 61–74 | 67

Scheme 3 TN, T and Sialyl-T MUC1 Pam3Cys two-component vaccines.
prepared by Fmoc solid-phase synthesis having the amino acid side chain
protecting groups still protected after resin cleavage. Different TN, T and
sialyl T MUC1 glycopeptides (37–39) were further prepared with both the
glycan and peptide protecting groups removed. Peptide synthesis was fol-
lowed by fragment condensation employing HATU and HOAt and after
additional deprotection resulting in the formation of the lipopeptide vaccine
constructs 41–43. Immunization of these vaccines in mice showed that a
specific immune response was induced in all mice although not with the
extraordinary high antibody titers found employing MUC1 Tetanus Toxoid
vaccines.5f In a later study di- and tri component vaccines were synthesized
and compared by immunological evaluation. The three-component vac-
cines, containing an extra T-cell peptide epitope, showed here a stronger
immune response if the MUC1 peptides were glycosylated while the non-
glycosylated two- and three-component vaccines did not show any differ-
ence in antibody titers.5e
4 Dendrimer vaccines
Like the lipopeptide Pam3CSK4 vaccines, constructs based on dendrimer
multi-epitope presentation can avoid that carrier-induced immune sup-
pression take place during immunization. Vaccines based on multiple
antigen presentation may further facilitate antigen uptake and presentation
by antigen presenting cells (APCs). Linear peptide backbones, cyclic peptide
scaffolds or multi-lysine scaffolds are commonly applied for multivalent
presentation of glycans or glycopeptides.8 Recently MUC1 and MUC4
TN- and sialyl-TN-glycopeptide dendrimer two- and three component vac-
cines based on the oligo-lysine core were reported.6a,9 A MUC1 sialyl-TN-
glycopeptide dendrimer vaccine was prepared either with or without
an extra immunostimulating Tetanus Toxoid T-cell peptide epitope. In
between the B- and T-cell epitope and the di-lysyl-lysine core a polar non-
immunogenic triethylene glycol spacer amino acid was incorporated. The
sialyl-TN-MUC1-glycopeptide 45 and the Tetanus Toxoid T-cell peptide
epitope 47 with a spacer at the C-terminal end were prepared by Fmoc-
solid-phase peptide synthesis. Starting with aminomethyl-polystyrene resin
(AMPS) functionalized with an allylic HYCRON linker the assembly of
the peptide backbone were performed according to standard procedures.
Palladium(0)-catalyzed cleavage of the allylic linker then gave the MUC1
glycopeptide 45 and the Tetanus Toxoid peptide 47. For construction of
the dendrimers, fragment condensation of sialyl-TN-glycopeptide 45 to the
di-lysyl lysine core 44 was performed employing HATU/HOAt and N-
methyl morpholine for activation and resulted in the MUC1-glycopeptide
dendrimer. Acidolytic removal of the peptide side-chain protecting groups
and glycan deprotection by catalytic hydogenation of the benzyl ester
and careful deacetylation using catalytic sodium metoxide in methanol
generated the dendrimer vaccine 46. In a similar fashion a glycopeptide
dendrimer vaccine including a Tetanus Toxoid T-cell epitope was pre-
pared. Fragment condensation of the Tetanus Toxoid peptide 47 to the
solid-phase-linked di-lysine core 44 was carried out. The N-terminal
di-benzyloxycarbonyl protecting group was then removed by
68 | Carbohydr. Chem., 2012, 38, 61–74

hydrogenation employing palladium on charcoal, followed by coupling of
the carboxy-activated MUC1-glycopeptide 45. Subsequent removal of the
peptide and glycan protecting groups gave the dendrimer vaccine 48. The
prepared synthetic dendrimer vaccines are currently under investigation in
immunization studies6a (Scheme 4).
Self-adjuvanting multivalent MUC1 T and TN glycopeptide vaccine
constructs efficiently conjugated through azide alkyne click chemistry to the
Pam3CSK4 lipopetide immune-stimulant have also been prepared.6b On
solid-phase, the Pam3CSK4 lipopeptide was prepared with an extra
C-terminal lysine residue with the side-chain protected with a 1-(4,4-
dimethyl-2,6-dioxocyclohexylidene)-3-methyl-butyl (ivDde) group, which
could be selectively removed using hydrazine monohydrate. Repeated
coupling with Fmoc-Lys(Fmoc)-OH resulted in a branched di-lysine core
structure which was alkyne functionalized by coupling with a triethyle-
neglycol-alkyne spacer. Glycopeptides 50–53 containing a N-terminal azide
spacer were prepared by standard Fmoc-solid-phase-synthesis. Through
copper (I) mediated click chemistry the lipopeptide alkyne construct 49 were
then efficiently coupled to glycopeptides 50–53 generating the multivalent
vaccine constructs 54–57. These vaccines are still to be tested for potential
immunogenicity (Scheme 5).
5 Antibody epitope mapping

For development of safe anti-tumor vaccines, it is important that induced
antibodies are specific to the tumor-associated MUC1 glycopeptide struc-
tures and do not cross-react with glycan structures found on normal cells.
Variations between the tumor-associated glycan structures, the positions
that are glycosylated and glycosylation density per tandem repeat are seen
depending on tumor tissue and stage of disease (stage I-IV). Correlation
between antibody binding epitope analysis and binding preferences to dif-
ferent tumor tissue or tumor cell-lines, contribute to conclude which
structures that are relevant for future vaccine design. Studies of antibody
specificity are also important in order to evaluate immune response differ-
ences between individuals. Through antibody neutralization experiments or
by immobilization of glycopeptides on microarrays the binding preferences
of antibodies induced with the synthetic MUC1 vaccines were further stu-
died.2a,5a,b Antibodies induced with vaccines containing sialyl-TN glycosy-
lation on Thr-6 on the MUC1 glyccopeptide conjugated to T.Tox. protein
or to a OVA T-cell epitope, were highly specific to both the sialyl-TN glycan
and to the HGVT glycosylation position. If the sialyl-TN glycan was moved
to a different position on the repeat, then the peptide was not recognized by
the antibodies. Glycosylation with TN, or 2,6-sialyl-T in the HGVT position
resulted in moderate binding and glycosylation with T, 2,3-sialyl-T or
no-glycosylation resulted in very weak or loss of binding recognition.
Evaluation of antibodies induced with T.Tox. conjugated to MUC1 gly-
copeptides with the T-antigen or a 2,6-di-fluoro-T antigen in Thr-6 in the
HGVT region, resulted in completely different specificity pattern. These
antibodies showed a broad binding specificity, and the sialyl-TN, TN, T,
sialyl-T and the non-glycosylated MUC1 tandem repeat peptides were all
Carbohydr. Chem., 2012, 38, 61–74 | 69

NH2 O NHR O
(a) H H
N N
H 2N O-t-Bu RHN OH
O 1. TFA
HATU, HOAt, 2. Pd(C), H2, MeOH O
HO OH CO2H
HO DMF, STNMUC1 (45) 3. NaOMe/MeOH
O NH2 NHR
AcHN O NH
HO HO H2N NH
RHN
O O O
HO 44 46
AcHN O HO
46, R = O O O O O H O
H O H H H
N N N N N N N
N N N N N N O
H O H O H O O H O H O O 3
OH O
HO
NH
H 2N NH
NH2 O
H NH2R
H O
N
H 2N O-t-Bu N
(b) R2HN OH
O 1. Pd(C), H2, MeOH
HATU, HOAt, 1. TFA O
2. HATU, HOAt,
2. Pd(C), H2, MeOH
HO NH2 DMF, T.Tox. pep.(47) DMF, ST MUC1 (45)
OH CO2H 44 N 3. NaOMe/MeOH NHR2
70 | Carbohydr. Chem., 2012, 38, 61–74

HO NH
O H2N NH
AcHN O
R2HN
O
HO HO 48 O
O
HO
AcHN O HO
O H O H O H O H O O
N N N N N N O
N N N N N N
H O H O H O O H O H O HN
OH O
O
HO NH
H2 N NH HO O
HO
48, R2 = O
HO
O H O O O
H H
O N N N
N N N N N
O 3 H O H H H O
O O
OH
Scheme 4 Synthesis of a) sialyl-TN-MUC1-glycopeptide dendrimer, b) sialyl-TN-MUC1-T.Tox dendrimer vaccine.

OH OH OH HO OH
HO
TN = T= O O
O HO O
H2N NH HO
MUC1 = HO AcHN
AcHN
NH
HO
OH O HO
O O O O O O O O O O O
H H H H H H H
N N N N N N N N OH
N N N N N N N N N N N
H H H H H H H
O O O O O O O O
HO O O
N NH
R2 R1 50, 54: R1 = TN, R2 = H
51, 55: R1 = TN, R2 = TN
O 52, 56: R1 = T, R2 = H
O 53, 57: R1 = TN, R2 = T
Pam3CSK4 + N3 MUC1
O
H
O N N
N O
50–53 H N MUC1
49 N O 3
CuI O
3
O
O
H
HN N O N
NH2 NH2 N MUC1
N
HN O 3
O O
3 H
HO O N O N
O O O O N MUC1
O C15H31
H H H H H N
N N N O 3
N N N O
S N N N H 3
H H H O O
O HN O O O O
O HO H
C15H31 C15H31 N O N
O N
N MUC1
NH2 HN O 3
NH2 O
3
Toll-Like receptor Ligand Pam3CSK4 54–57 O
Carbohydr. Chem., 2012, 38, 61–74 | 71

Scheme 5 Multivalent MUC1-Pam3CSK4 dendrimer vaccines obtained by copper(I)-mediated click reactions.
recognized by the induced antibody sera. The MUC1 peptide backbone
seemed to be particularly important for antibody binding and truncation of
the peptide backbone to a 12-mer peptide resulted in that the binding
drastically dropped. The MUC1 T-antigen could further not be neutralized
by a glycopeptide from a different mucin, MUC4. In contrast to the more
specific antibodies initially induced, the antibodies induced with the
T-antigen or 2,6-di-fluoro-T antigen MUC1 T.Tox. vaccines with a slightly
broader binding recognition did strongly recognize breast cancer tumor
cells employing FACS analysis.
Analysis of monoclonal antibodies known to recognize tumor tissue
or analysis of auto-antibodies induced in patients with different stages of
cancer, may also be valuable in vaccine design. The SM3 monoclonal
antibody induced from deglycosylated mucin from human milk, known
to bind to tumor tissue, was studied in detail.2a,10 Through microarray
glycopeptide analysis, it was found that the antibody do not only bind to
the non-glycosylated MUC1 tandem repeat or MUC1 glycosylated with
TN in the immunodominant PDTR region previously found, but glycosy-
lation with TN or sialyl-TN in the GSTA region was also strongly
recognized. In later studies employing glycopeptide microarrays, human
serum auto-antibodies from cancer patients were found to be directed to
MUC1 glycopeptide epitopes glycosylated with sialyl-TN on the Thr or Ser
in the GSTA region.2b These findings resulted in the preparation of T.Tox
sialyl-TN MUC1 glycopeptide vaccines glycosylated in these positions
(9, 10). As described above, antibodies induced from these vaccines (9, 10)
recognize both cancer cells from the MCF-7 cell-line and breast cancer
tissue.5c
Interestingly, in another study sera were analysed from 395 patients with
early stage breast cancer.11 High levels of serum antibodies against the
sialyl-TN MUC1 and also against core 3 glycans were found, which could be
associated with a benefit in metastasis free survival. The core 3 glycan and
the glycosyltransferase (b3GNT6) responsible for core 3 glycan extension, is
however not found in normal or malignant breast tissue. Core 3 antibody
induction may therefore be a result of mimicry or due to inflammation in
the colon, where these glycans commonly are expressed.12 In summary it
could be concluded that sialyl-TN glycosylation on MUC1 tandem repeats is
a highly relevant breast cancer target both as diagnostic marker and for
therapeutic applications.
6 Conclusions
During the recent years, a number of vaccines consisting of tumor-asso-
ciated MUC1-glycopeptides containing the TN-, T-, sialyl-TN- and sialyl-T-
saccharide antigens have been synthesized by their conjugation to different
immuno-stimulants. Immunological evaluation of these vaccines has shown
that antibodies were induced which are specific towards tumor-associated
glycopeptides and do recognize the membrane glycoproteins on cancer cells.
Several of these synthetic vaccines showed promising effects, in particular
the MUC1 Tetanus Toxoid vaccines. Furthermore, extensive synthesis of
glycopeptides for antibody analysis in a microarray format make it possible
72 | Carbohydr. Chem., 2012, 38, 61–74

to identify new potential vaccine candidates as well as to study in detail the
antibody binding epitopes induced by the different vaccines.
References
1 (a) G. F. Springer, Science 1984, 224, 1198–206; (b) J. Taylor-Papadimitriou,
J. Burchell, D. W. Miles, M. Dalziel, Biochim. Biophys. Acta, Mol. Basis Dis.
1999, 1455, 301–313; (c) T. Becker, S. Dziadek, S. Wittrock, H. Kunz, Curr.
Cancer Drug Targets 2006, 6, 491–517.
2 (a) U. Westerlind, H. Schroeder, A. Hobel, N. Gaidzik, A. Kaiser, C. M.
Niemeyer, E. Schmitt, H. Waldmann, H. Kunz, Angew. Chem. Int. Ed. 2009, 48,
8263–8267; (b) O. Blixt, E. Clo, E. Nudelman, K. Kildegaard Sö rensen,
T. Clausen, H. H. Wandall, P. Livingston, H. Clausen, K. J. Jensen, J Proteome
Res. 2010, 9, 5250–5261; (c) H. H. Wandall, O. Blixt, M. A. Tarp, J. W.
Pedersen, E. P. Bennett, U. Mandel, G. Ragupathi, P. Livingston, M. A.
Hollingsworth, J. Taylor-Papadimitriou, J. Burchell, H. Clausen, Cancer Res
2010, 70, 1306–1313.
3 (a) I. Brockhausen, Biochim. Biophys. Acta, Gen. Subj 1999, 1473, 67–95;
(b) I. Brockhausen, J.-M. Yang, J. Bruchell, C. Whitehouse, J. Taylor-Papa-
dimitriou, Eur. J. Biochem. 1995, 233, 607–17; (c) J. M. Burchell, A. Mungul, J.
Taylor-Papadimitriou, J Mammary Gland Biol 2001, 6, 355–64; (d) K. O. Lloyd,
J. Burchell, V. Kudryashov, B. W. T. Yin, J. Taylor-Papadimitriou, J. Biol.
Chem. 1996, 271, 33325–33334.
4 (a) B. J. Swanson, K. M. McDermott, P. K. Singh, J. P. Eggers, P. R. Crocker,
M. A. Hollingsworth, Cancer Res. 2007, 67, 10222–10229; (b) L. G. Yu,
N. Andrews, Q. Zhao, D. McKean, J. F. Williams, L. J. Connor, O. V.
Gerasimenko, J. Hilkens, J. Hirabayashi, Kasai, K.-i., J Biol Chem 2007, 282,
773–81.
5 (a) A. Kaiser, N. Gaidzik, U. Westerlind, D. Kowalczyk, A. Hobel, E. Schmitt,
H. Kunz, Angew. Chem. Int. Ed. 2009, 48, 7551–7555; (b) A. Hoffmann-
Roeder, A. Kaiser, S. Wagner, N. Gaidzik, D. Kowalczyk, U. Westerlind, B.
Gerlitzki, E. Schmitt, H. Kunz, Angew. Chem. Int. Ed. 2010, 49, 8498–8503; (c)
N. Gaidzik, A. Kaiser, D. Kowalczyk, U. Westerlind, B. Gerlitzki, H. P. Sinn,
E. Schmitt, H. Kunz, Angew. Chem. Int. Ed. 2011, 50, 9977–9981; (d) H. Cai,
Z.-H. Huang, L. Shi, Z.-Y. Sun, Y.-F. Zhao, H. Kunz, Y.-M. Li, Angew. Chem.
Int. Ed. 2012, 51, 1719–1723; (e) B. L. Wilkinson, S. Day, L. R. Malins, V.
Apostolopoulos, R. J. Payne, Angew. Chem. Int. Ed. 2011, 50, 1635–1639; (f)
A. Kaiser, N. Gaidzik, T. Becker, C. Menge, K. Groh, H. Cai, Y.-M. Li,
B. Gerlitzki, E. Schmitt, H. Kunz, Angew. Chem. Int. Ed. 2010, 49, 3688–3692;
(g) V. Lakshiminarayanan, P. Thompson, M. A. Wolfert, T. Buskas, J. M.
Bradley, L. B. Pathangey, C. S. Madsen, P. A. Cohen, S. J. Gendler, G. J.
Boons, Proc. Natl. Acad. Sci. 2012, 109, 261–266.
6 (a) S. Keil, A. Kaiser, F. Syed, H. Kunz, Synthesis 2009, (8), 1355–1369; (b) H.
Cai, Z.-H. Huang, L. Shi, Y.-F. Zhao, H. Kunz, Y.-M. Li, Chem. Eur. J. 2011,
17, 6396–6406.
7 H. Cai, Z.-H. Huang, L. Shi, P. Zou, Y.-F. Zhao, H. Kunz and Y.-M. Li, Eur J
Org Chem, 2011, 3685–3689.
8 (a) A. Schierholt, M. Hartmann, T. K. Lindthorst, Carbohyd. Res. 2011, 346,
1519–1526; (b) I. Bossu, M. Sulc, K. Krenek, E. Dufour, J. Garcia, N. Berthet,
P. Dumy, V. Kren, O. Renaudet, Org. Biomol. Chem. 2011, 9, 1948–1959;
(c) E. M. V. Johansson, S. A. Crusz, E. Kolomiets, L. Buts, R. U. Kadam, M.
Cacciarini, K.-M. Bartels, S. P. Diggle, M. Camara, P. Williams, R. Loris, C.
Nativi, F. Rosenau, K.-E. Jaeger, T. Darbre, J.-L. Reymond, Chemistry &
Carbohydr. Chem., 2012, 38, 61–74 | 73

biology 2008, 15, 1249–1257; (d) X. Tian, J. Pai, K.-H. Baek, S.-K. Ko, I. Shin,
Chem. Asian. J. 2011, 6, 2107–2113.
9 (a) J. P. Tam, Proc. Natl. Acad. Sci. 1988, 85, 5409–13; (b) T. Becker, A. Kaiser,
H. Kunz, Synthesis 2009, (7), 1113–1122.
10 (a) J. Burchell, S. Gendler, J. Taylor-Papadimitriou, A. Girling, A. Lewis, R.
Millis, D. Lamport, Cancer Res. 1987, 47, 5476–82; (b) J. Burchell, J. Taylor-
Papadimitriou, M. Boshell, S. Gendler, T. Duhig, Int. J. Cancer 1989, 44,
691–6.
11 O. Blixt, D. Bueti, B. Burford, D. Allen, S. Julien, M. A. Hollingsworth, A.
Gammerman, I. Fentiman, J. Taylor-Papadimitriou, J. Burchell, Breast Cancer
Res 2011, 13, R25.
12 (a) B. J. Campell, L. G. Yu, J. M. Rhodes, Glycoconj 2001, 18, 851–858; (b)
J. W. Pedersen, O. Blixt, E. P. Bennett, M. A. Tarp, I. Dar, U. Mandel, S. S.
Poulsen, A. E. Pedersen, S. Rasmussen, P. Jess, H. Clausen, H. H. Wandall, Int
J Cancer 2010, 128, 1860–1871.
74 | Carbohydr. Chem., 2012, 38, 61–74

Posttranslational sialylation and its impact
on leukocyte recruitment during
inflammation
Karin Bodewits and Markus Sperandio*
DOI: 10.1039/9781849734769-00075
The immune response to invading pathogens during inflammation is crucial

for viability of humans and other mammals. It needs leukocytes to leave
the blood vessel system (intravascular compartment) and transmigrate
into infected or inflamed tissue to fulfil its function as immune cells. The
navigation of immune cells during this process requires a complex network
of adhesion, signalling and direction molecules on the immune cells and
the blood vessel wall. These molecules include selectins, selectin-ligands,
chemokines, chemokine receptors, and integrins. Carbohydrate decorations
of some of these adhesion and signalling molecules play a major role
in effective leukocyte recruitment, as they can regulate proper function,
stability, protease resistance, or quaternary structure. In this chapter, the
role of carbohydrate decorations in leukocyte recruitment with a strong
focus on posttranslational modification by sialic acids will be reviewed.
1 Leukocyte recruitment cascade

Leukocytes continuously circulate through the vasculature, from one tissue
to another and to acute and chronic sites of inflammation. Leukocyte
trafficking to tissues where leukocytes are required serves a critical function
in the mammalian immune response and is mediated by a large number
of adhesion and signalling molecules. The first reports on leukocyte
recruitment date back to the early nineteenth century [1, 2]. However, only
during the last two to three decades, the molecular mechanisms governing
leukocyte recruitment have been elucidated. We have learned that during
acute inflammation the recruitment process of mostly neutrophils primarily
occurs in postcapillary venules of the affected tissue and follows a
well-characterised cascade of events that takes place in the presence of
hemodynamic shear forces exerted by the flowing blood (Fig. 1) [3–5].
Neutrophils, also known as polymorphonuclear (PMN) leukocytes, are
the most abundant leukocytes in the peripheral blood of humans and are
rapidly and efficiently recruited to areas of inflammation. The recruitment
process starts with the capture of free flowing neutrophils to the vessel wall,
followed by rolling along the endothelium. Capture and rolling of the cells
along the endothelium is mediated by a group of glycoproteins, called
selectins, which interact with carbohydrate determinants on selectin ligands
[6–8]. During rolling, neutrophils slow down and get into close contact with
endothelial cells, which enable endothelial-bound chemokines to interact
with specific chemokine receptors present on the cell surface of neutrophils.
Walter-Brendel-Center of Experimental Medicine, Ludwig-Maximilians-Universität,

Marchioninistr. 15, 81377 Munich, Germany. E-mail: markus.sperandio@lmu.de
Carbohydr. Chem., 2012, 38, 75–93 | 75

Fig. 1 The multistep leukocyte recruitment cascade. During inflammation, neutrophils or other
leukocytes are captured to the luminal surface of inflamed postcapillary venules, where they
start rolling along the endothelium. During rolling, chemokine/chemokine receptor interactions
trigger the activation of leukocyte-expressed integrins leading to firm arrest of leukocytes on the
inflamed endothelium followed by adhesion strengthening and spreading on the endothelium.
To find a suitable way out of the vasculature, leukocytes start crawling along the endothelium
until the right spot has been found for transmigration.
This in turn triggers the activation of leukocyte-expressed integrins

(inside-out signalling), which interact with endothelial-expressed ligands,
leading to firm adhesion and ultimately to the arrest of neutrophils on the
endothelium [9, 10]. Thereafter, adhesion is further strengthened and
neutrophil spreading induced by integrin/integrin ligand interactions
(outside-in signalling) [11]. Interestingly, neutrophils do not readily trans-
migrate through the vessel wall into tissue after the initial firm arrest, but
start to slowly crawl on the endothelial luminal surface in search of a good
exit point into the inflamed or infected tissue [12]. Upon entering the tissue,
gradients of chemokines and other chemoattractants (chemotaxis) direct the
further migration of the cells [13].
1.1 Leukocyte capture and rolling: selectins and their ligands

The capture of free flowing leukocytes and rolling along the endothelium is
considered a crucial step during the recruitment of leukocytes into inflamed
tissue. Leukocyte capture and rolling are mediated by selectins recognising
and binding to specific carbohydrate determinants on selectin ligands [14].
Three different types of selectins have been identified in mammals: leukocyte
(L)-, platelet (P)-, and endothelium (E)-selectin [8]. L-selectin is con-
stitutively expressed on the surface of most leukocytes and can be shed from
the cell surface upon stimulation. P-selectin is stored in Weibel-Palade
bodies of endothelial cells and is rapidly mobilised to the cell surface during
inflammation or injury. As the name implies, P-selectin is also expressed by
activated platelets. Like P-selectin, E-selectin is expressed on endothelial
cells. However, E-selectin does not reside in intracellular storage pools, but
is newly synthesised after stimulation by different inflammatory mediators.
In the mouse, the predominant leukocyte rolling receptor in acutely
76 | Carbohydr. Chem., 2012, 38, 75–93

inflamed endothelial cells is P-selectin. However, this might be different in
humans [15]. P-selectin is specialised in capturing free circulating leukocytes
from the blood stream, whereas E-selectin is more liable for stabilising the
rolling movement and slowing it down [16].
All three selectins are members of a glycoprotein family called C-type
lectins [17], which means they recognise and bind to specific carbohydrate
determinants (lectins) and typically need calcium (C-type) for binding. The
best-known binding ligand of selectins is the tetrasaccharide sialyl-Lewis-X
(sLeX), which will be discussed later in this chapter. Endothelium-expressed
E- and P-selectin bind to selectin ligands expressed on rolling leukocytes.
The selectin ligands comprise a heterogeneous group of proteins and
glycolipids that are all decorated with specific carbohydrate determinants
containing sialic acid modifications. Most of the protein ligands bear the
selectin-binding carbohydrate moiety at the N-terminus of the protein away
from the cell membrane [18]. P-selectin glycoprotein ligand-1 (PSGL-1), the
most prominent member, is expressed on the surface of most leukocytes and
functions as a ligand for P-, E- and L- selectin [19, 20]. Besides PSGL-1,
many other leukocyte-expressed selectin ligands have been described,
among them, the E-selectin ligands CD44, E-selectin ligand-1 (ESL-1), and
CD43 [21, 22].
After capturing, neutrophils start to roll on the activated endothelium.
The rolling velocity on the vessel wall is B100–1000-fold slower than the
average blood flow velocity [14]. In general, for leukocytes to roll against
hemodynamic shear forces, selectin binding to its ligand needs to comply
with the following three requirements: 1) rapid bond formation, 2) high
tensile strength during binding, and 3) fast dissociation rates [23]. In the case
of selectins and their ligands, we know that bond formation requires the
presence of shear forces to support rolling. In addition, it has been shown
for L-selectin dependent rolling that leukocytes detach from the endothe-
lium when blood flow velocity falls under a critical velocity level [24].
Furthermore, under physiological conditions, bonds become more resistant
to shear forces with increasing shear stress (catch bonds) [25]. This
characteristic behaviour of selectins is of utmost importance for the
recruitment of free flowing leukocytes to the vessel wall and enables sub-
sequent adhesion and transmigration of leukocytes [8].
1.2 Leukocyte adhesion: chemokines, integrins, and their ligands

Before transendothelial migration takes place, firm adhesion of leukocytes
to the vessel wall occurs, and leukocytes stop rolling. The transition from
rolling to arrest can be triggered by chemokine-or selectin-dependent sig-
nalling events leading to the activation of leukocyte-expressed integrins [10].
Chemokines are a large family of small proteins, 8- to 12-kD in size. Some
of its members including CXCL1 (keratinocyte-derived chemokine, KC) or
CXCL8 (IL-8) can be found on the luminal surface (glycocalyx) of inflamed
endothelium in response to pro-inflammatory stimuli [26, 27]. During
rolling, chemokine receptors expressed on neutrophils (including CXCR2)
sense the repertoire of endothelium-immobilised chemokines on the luminal
surface of the vessel wall. Specific interactions between chemokines and
their respective chemokine receptors then induce intracellular signalling
Carbohydr. Chem., 2012, 38, 75–93 | 77

events (inside-out signalling), leading to leukocyte integrin activation and
firm leukocyte arrest on the endothelium [4].
Integrins are type I transmembrane glycoproteins, composed of an a and
a b subunit, and are amongst others expressed on the surface of leukocytes.
Integrins that play an important role for neutrophil recruitment include the
b2 integrins leukocyte function-associated antigen 1 (LFA-1, aLb2) and
macrophage-1 antigen (Mac-1, aMb2). b2-integrins are required for slow
neutrophil rolling, adhesion strengthening, spreading, crawling towards
transmigration sites, and transmigration (diapedesis) through the vessel wall
[28]. b2 integrins are constitutively expressed on free flowing leukocytes, but
are mostly kept in their inactive form awaiting activation. They can have
three different conformations of their extracellular domain: the bent or
inactive form that prohibits binding to its receptors, the extended inter-
mediate-affinity form, and the extended high-affinity conformation [29].
Chemokine-triggered activation causes a rapid conformational change from
the inactive (bent) to the extended intermediate-affinity form. This change
allows the integrin to bind to its endothelial-expressed ligand, leading to a
further conformational shift to the extended high-affinity conformation.
This full repertoire of integrin activation is required for leukocyte firm
arrest on the vessel wall [28].
Relevant endothelium-expressed integrin ligands for LFA-1 include
intercellular adhesion molecules-1 and -2 (ICAM-1, ICAM-2) [30, 31].
Mac-1 binds to the adhesion molecules ICAM-1 and the endothelial
receptor for advanced glycation end products (RAGE) [32].
After integrin mediated firm arrest on the vessel wall, neutrophils
undergo several morphological changes including adhesion strengthening
and spreading which prepare the neutrophil to slowly move (crawling) to an
appropriate location where it can exit the vasculature [12].
1.3 Transendothelial migration (diapedesis)

The attachment of leukocytes to the vessel wall and transmigration to
inflamed tissue was described more than 150 years ago, and much progress
in understanding the diapedesis step has been made since by using electron
microscopy (EM), which enabled the observation of morphological changes
of both the leukocyte and the endothelium during diapedesis. Two modes of
transmigration have been described: the paracellular route and the trans-
cellular route [33]. The paracellular route, which is the most widely accepted
route, describes the migration of leukocytes between the junctions of two or
more endothelial cells. The transcellular route describes the migration of
leukocytes through endothelial cells [34]. The understanding of the mole-
cular mechanism of both diapedesis routes has only just begun to emerge
from studies conducted in genetically engineered mice with the identification
of different adhesion and signalling molecules (including ICAM-2,
PECAM-1, JEM-C, CD99, VE- cadherin) to be involved in transmigration.
2 Posttranslational modifications
The cells of all organisms are composed of four fundamental macro-
molecular components: nucleic acids, proteins, glycans, and lipids [35].
78 | Carbohydr. Chem., 2012, 38, 75–93

The term glycan comprises any form of mono-, oligo- or polysaccharide,
which can exist in its unbound form or connected to another molecule.
Glycans differ from the amino acids and nucleic acid polymers in three
ways: 1) they are built-up out of a large repertoire of monomers, 2) they can
be heavily branched structures, and 3) the linkage types between the dif-
ferent monomeric subunits are highly variable [36]. The glycome of a single
species can consist of hundreds of thousands of distinctive carbohydrate
structures, and can by all its diversity carry much more information than the
other three building blocks of life [37, 38]. Glycosylation is ubiquitous
among eukaryotes and though the historical perception of glycans was that
they are mere cellular decorations, we now know that they are of central
importance in a wide variety of biological processes, including host-
pathogen interactions, cellular trafficking, and intercellular adhesion [39].
Glycosylation can significantly contribute to the biological properties of
proteins and lipids including its conformation, solubility, or function as
selective ligand. In this context, the influence of carbohydrate decoration on
immune function and leukocyte trafficking has gained increasing interest in
recent years. Selectin-selectin ligand interactions are among the best-studied
carbohydrate-protein interactions known in biology [18, 40]. Furthermore,
several studies have shown that posttranslational modifications are
functionally relevant for chemokine-dependent arrest of the leukocytes on
the vessel wall and integrin-mediated leukocyte adhesion [41, 42].
2.1 Glycans
Glycans decorate the surfaces of cells and are attached to lipids and
proteins. Despite their diversity, each class of glycans has a limited number
of core structures, and the different capping (outer) carbohydrate groups
create the high overall diversity. The terminals, or outer capping carbo-
hydrates, are also often responsible for the biological functions of the
carbohydrate decoration [43].
Carbohydrate decorations of cell surface- and secreted proteins are
typically covalently attached to proteins and lipids through either a
nitrogen atom (N-glycans) or an oxygen atom (O-glycans). The nitrogen
atom is supplied by the amino acid asparagine for N-glycans and either a
serine or threonine residue of the protein can supply the oxygen atom as
attachment point for O-glycans [36]. The different sugar monomers are
attached to the underlying protein or glycoconjugate by enzymes called
glycosyltransferases.
Selectin ligands carry simple and complex sialylated and fucosylated
sequences, in some ligands sulphated sequences can be found as well. A
common carbohydrate structure found on selectin ligands is sialyl Lewis X
(sLeX). sLeX is a tetrasaccharide consisting of galactose (Gal), fucose (Fuc),
N-acetylglucosamine (GlcNAc), and sialic acids (Fig. 2). It is biosynthesised
by several glycosyltransferases in the Golgi complex and can be linked to
a diverse variety of glycoconjugates, such as N-glycans, O-glycans, and
glycolipids. On selectin ligands, sLeX is typically found as a capping group
of core 2-decorated O-glycans and is crucial for binding to selectins. Some
endothelial selectin ligands which are important in lymphocyte homing
carry a sulfated form of sLex, 6-sulfo-sLeX, which is generated by the help of
Carbohydr. Chem., 2012, 38, 75–93 | 79

Fig. 2 Key carbohydrate decoration of selectin ligands. The tetrasaccharide sialyl Lewis X
(sLeX) is shown as the capping group (bioactive group) of core 2 decorated O-glycans attached
to a serine or threonine residue. The presence of sLeX on the outside of selectin ligands is critical
for the recognition by L-, P-, and E-selectin during leukocyte recruitment. The GlcNAc can be
sulfated at C6-position, forming 6-sulfo-sLex.
two additional GlcNAc6-sulfotransferases called GlcNAc6ST-1 and

GlcNAc6ST-2 [44].
2.2 Glycosyltransferases involved in leukocyte trafficking

The posttranslational modification of proteins and lipids is a dynamic, con-
trolled, and consecutive process. Carbohydrate chains are biosynthesised or
biodegraded in a well-organised mode due to the different substrate specificities
of glycosyltransferase and glycosidase enzymes [45]. In eukaryotes, post-
translational glycosylation mainly takes place in the Golgi apparatus and is
dependent on a group of Golgi resident enzymes named glycosyltransferases
[46]. Glycosyltransferases form the predominant integral membrane proteins
of the Golgi-machinery membranes and over 200 have been identified in the
human genome [40]. The sugar substrates of the glycosyltransferases can be
biosynthesised and activated in different compartments of the cell, such as
cytoplasm and nucleus, before being transferred to the Golgi apparatus.
Most glycosyltransferases belong to the type II transmembrane proteins
that have a cytoplasmic tail (CT), a transmembrane-spanning domain
(TM), and a stem plus catalytic domain. They have been identified as
monomers, homo-dimers, hetero-dimers, and hetero-oligomers [47]. These
enzymes specifically transfer activated sugar nucleotide donors to a specific
OH-group of a glycoconjugate via a a- or b- linkage. Glycosyltransferase
activity can be regulated at several levels, which include transcriptional
80 | Carbohydr. Chem., 2012, 38, 75–93

regulation of the genes encoding the transferases or activation of the
enzymes by the presence of their substrates [45].
Several glycosyltransferases involved in the biosynthesis of functional
adhesion and signalling molecules essential for immune cell trafficking have
been described, including glucosaminyltransferases, galactosyltransferases,
fucosyltransferases, and sialyltransferases [18]. The glycosyltransferases
involved in selectin ligand function and leukocyte recruitment are summarised
in Table 1 and will be discussed briefly in the following subparagraphs.
Table 1 Glycosyltransferases and their contribution to leukocyte recruitment in vivo.
Gene name Phenotype in the knock-

Glycosyltransferases (mouse) Function out mouse
Selectin ligand function

Glucosaminyltransferases
’ C2GcT-1 Gcnt1 Extension of the core 2 Impaired P-, L-, and
O-glycan branch. Adds E-selectin mediated roll-
GlcNAc to GalNAc. ing of leukocytes along
the vessel wall [48–50]
Galactosyltransferases
’ b1,4-GalT1 B4galt1 Involved in synthesis of Leukocyte recruitment
polylactosamine. defect in several models
of acute and chronic
inflammation, impaired
P-selectin binding to
neutrophils [51].
Fucosyltransferases
’ FucT-4 Fut7 Transfers GDP-fucose to Increased E-selectin-
acceptor carbohydrates mediated leukocyte
within the glycan chain. rolling velocities in
inflamed venules [59].
’ FucT-7 Fut4 Transfers GDP-fucose to Higher systemic blood
terminal GlcNAc after leukocyte counts,
sialylation of the term- P- and E- selectin
inal galactose. mediated rolling is
clearly reduced [58].
(Fut4/Fut7 double knock-
out with almost com-
plete loss of leukocyte
rolling [57, 59])
Sialyltransferases
’ ST3Gal-IV St3gal4 Formation of selectin Reduced E-selectin
ligands (sLex). mediated rolling.
Increased E-selectin
dependent rolling
velocity [61]. Reduced L-
selectin dependent rolling
during inflammation [62].
Chemokine receptor
function
’ ST3Gal-IV St3gal4 Putative sialylation of Reduced CXCR2-trig-
CXCR2 gered neutrophil arrest
in vivo and in vitro [41].
Integrin activation
’ ST6Gal-I St6gal1 Sialylation of VLA4’s VLA4 activation [42].
extracellular domain
Carbohydr. Chem., 2012, 38, 75–93 | 81

2.2.1 Glucosaminyltransferases. A well-known glucosaminyltransferase
relevant for the synthesis of functional selectin ligands is core 2 b1,6N-
acetylglucosaminyltransferase (core2 GlcNAcT-1, gene name Gcnt1).
Core2GlcNAcT-1 is one of three glycosyltransferases that build the core 2
backbone and initiate the core 2 extension by adding a N-acetylglucosamine
(GlcNAc) to the serine or threonine-bound N-acetylgalactosamine (GalNAc)
via a b1,6 linkage (Fig. 2). In Gcnt1-deficient mice, P- and L- selectin-mediated
leukocyte rolling during inflammation in vivo is strongly attenuated, while
E-selectin dependent rolling is only moderately affected [48–50].
2.2.2 Galactosyltransferase. As Fig. 2 depicts, b-1,4-galactosyl-
transferase activity is required to form the polylactosamine (GlcNAc-Gal)n
backbone which is found on selectin ligands. b1,4-GalT-I links galactose to
terminal GlcNAc residues via a b1,4 linkage and is involved in the synthesis
of the terminal capping group sLeX [51]. B4galt1 deficient mice showed
impaired leukocyte recruitment in several models of acute and chronic
inflammation [51]. In addition, B4galt1 deficient neutrophils exhibited
reduced binding to soluble P-selectin. Of note, recently B4GALT1 deficient
patients have been identified. The disease has been termed CDG IId
(congenital deficiency of glycosylation-IId), but no severe inflammatory
phenotype has been described so far [52, 53].
2.2.3 Fucosyltransferases. Fucosyltransferases transfer fucose from the
donor GDP-fucose to acceptor molecules which can be oligosaccharides,
glycoproteins or glycolipids. Two fucosyltransferases are important for the
biosynthesis of selectin ligands: FucT-4 and FucT-7 [54, 55]. Both enzymes
are expressed in the Golgi apparatus of leukocytes. In blood vessel endo-
thelial cells, FucT-4 is only expressed in larger venules, whereas FucT-7
expression is mainly restricted to small postcapillary venules [56]. Either or
both genes encoding FucT-4 and FucT-7 [57–59], have been deleted in mice.
In Fut4-deficient mice the phenotype is rather mild with no defect in P- and
L-selectin-dependent rolling. However, E-selectin-dependent leukocyte
rolling velocities are significantly increased in inflamed venules, compared
to wild-type species, suggesting a defect in E-selectin ligand function [59].
On the contrary, the effects of deleting Fut7 are dramatic. This is already
illustrated by the significantly augmented number of circulating peripheral
leukocytes (a hallmark of defective leukocyte adhesion) in Fut7-deficient
mice [60]. In addition, intravital microscopy experiments revealed that
P- and E- selectin-mediated rolling are markedly reduced and L-selectin-
mediated lymphocyte homing severely impaired [58, 59]. In Fut4 and Fut7-
double deficient mice, P-, E-selectin, and L-selectin mediated rolling of
leukocytes are almost completely abolished, which clearly shows the crucial
role of fucosyltransferases in the biosynthesis of active selectin ligands [57].
2.2.4 Sialyltransferases. The sialyltransferases will be discussed in sec-
tion 3 of this chapter.
3 Sialic acids and sialyltransferases (ST)

Nine types of monosaccharides can be identified in the enzymatic process of
glycosylation in mammals. Well-conserved biosynthetic pathways for these
82 | Carbohydr. Chem., 2012, 38, 75–93

monosaccharides can be found in mammals that use sugars and other
precursors commonly present in the diet. In this section, the unique features
of sialic acids, the biosynthetic pathway, and the three main sialyl-
transferase families, will be discussed.
3.1 Sialic acids

Sialic acids are mainly found as glycoconjugate decorations on the cell
surface of eukaryotes and in some pathogenic microorganisms. The term
‘‘sialic acid’’ is commonly used to refer to the most abundant sialic acid
N-acetylneuraminic acid (Neu5Ac), although the sialic acids comprise a
family of nine-carbon sugars [63]. During evolution the sialic acid family
became greatly diversified and more than 50 different forms have been
isolated from a variety of natural sources [46, 64, 65]. The family comprises
neuraminic acid (Neu) derivatives with a shared nine-carbon backbone
(Fig. 3). They are a-keto acids and due to their net negative charge at
physiological pH, plus their bulky, hydrophilic chemical structure, they can
dramatically change the physical and chemical properties of the glycan
structures [66].
The basic structures of sialic acids carry either a N-acetyl group
(Neu5Ac) or a hydroxyl group at the 5-carbon, named 2-keto-3-deoxy-
nononic acid (KDN) (Fig. 3). From the biosynthetic perspective (discussed
in Section 3.1.1.), all sialic acids are likely to be derived from either Neu5Ac
or KDN [67]. They can be subject to a diverse range of modifications. A
common variation is the hydroxylation of the 5-N-acetyl group of Neu5Ac,
giving N-glycolylneuraminic acid (Neu5Gc), which is, together with
Neu5Ac and KDN, a representative member of the sialic acid family
(Fig. 3). The negatively charged carboxylate group at the 1-carbon can also
form a lactone ester with a hydroxyl group of a neighbouring sugar subunit.
Sometimes, the 5-carbon is deacetylated to give Neu. Furthermore, Neu,
Neu5Ac, KDN, and Neu5Gc can be further combined with O-acetyl,
Fig. 3 Natural diversity of sialic acids. The chemical structures of four common sialic acids
(e.g. Neu, Neu5AC, Neu5Gc, and Kdn) found in nature are shown. All sialic acids are believed
to be derived from either Neu5Ac or Kdn.
Carbohydr. Chem., 2012, 38, 75–93 | 83

O-lactyl, O-methyl, or O-sulfate groups at the different 4-, 7-, 8-, and 9-
carbon OH positions [68]. It has to be noted that in mice, only two forms of
sialic acids are found, namely Neu5Ac and Neu5Gc. In humans, only the
predominant form of sialic acids, Neu5Ac, has been discovered.
Sialic acids are not only complex sugars because of their high diversity
and net negative charge, but also because of their ability to form different
linkages to the underlying carbohydrate structure. The most common lin-
kages are a-2,3 and a-2,6 to galactose (Gal), a2,6 to N-acetylgalactosamine
(GalNAc), and a-2,8 linkage to neighbouring sialic acids. In addition, a-2,4
and a-2,9 linkages have been identified.
3.1.1 Sialic acid biosynthesis. As for the majority of sugars in nature the
basic building block of sialic acids is glucose. Glucose is converted in several
steps (not discussed here) into UDP-N-acetylglucosamine (UDP-GlcNAc),
which serves as the precursor of the sialic acid biosynthetic pathway. The
biosynthesis of sialic acid (Neu5Ac) involves four enzymatic steps after
the formation of UDP-GlcNAc (Fig. 4). The bifunctional enzyme GNE
Fig. 4 The Neu5Ac (sialic acid) biosynthesis pathway. After the formation of GlcNAc (derived
from glucose), the biosynthesis and activation of Neu5Ac involves five enzymatic steps. The
activated sialic acid, CMP-Neu5Ac, is utilised as a substrate by sialyltransferases in the Golgi
apparatus. ATP: adenosine triphosphate; CMP: cytidine monophosphate; CST: CMP sialic
acid transporter (also known as Slc35A1); CTP: cytidine triphosphate; GNE/MNK: UDP-
GlcNAc 2-epimerase/ManNAc-kinase; PEP: phosphoenolpyruvate; SAS: sialic acid phosphate
synthase; ST: sialyltransferases; UDP: uridine diphosphate.
84 | Carbohydr. Chem., 2012, 38, 75–93

(UDP-GlcNAc 2-epimerase)/ MNK (ManNAc-kinase) catalyses the first
two committed steps of the pathway. It firstly converts UDP-GlcNAc to
N-acetyl-mannosamine (ManNAc) with its UDP-GlcNAc-2-epimerase
activity. Thereafter, it catalyses the conversion of ManNAc to ManNAc-6-
phospate (ManNAc-6-P) using the kinase catalytic domain of the enzyme
[69, 70]. Subsequently, the sialic acid phosphate synthase (SAS) condenses
ManNAc-6-P with phosphoenolpyruvate, and Neu5Ac-9-phosphate is
formed [71]. Dephosphorylation of Neu5Ac-9-phosphate is catalysed by the
Neu5Ac-9-phosphate phosphatase (Neu5Ac-9-Pase) and Neu5Ac is gener-
ated [72].
Prior to incorporation of the sialic acid residues into carbohydrate dec-
oration of proteins, sialic acid has to be activated. In contrast to all other
vertebrate monosaccharaides that are activated with uridine- or guanosine
diphosphates (UDP or GDP, respectively), sialic acid is activated by the
cytidine monophosphate (CMP) [67]. The CMP-Neu5Ac synthetase that
condenses CMP and Neu5Ac to form CMP-Neu5Ac, is responsible for
activating sialic acids. Though all biosynthetic steps are taking place in the
cytosol, the activation of sialic acids into CMP-sialic acid takes place in the
cell nucleus [46]. After activation, CMP-sialic acid is transported from
the nucleus into the Golgi complex via a CMP-sialic acid transporter (CST,
also known as Slc35A1) [73], where it is utilised as a substrate of sialyl-
transferases. The amount of sialic acids produced by the cell is regulated by
a strong feedback inhibition mechanism of CMP-sialic acid acting on the
epimerase catalytic site of the GNE/MNK complex, catalysing the first step
of the biosynthesis [69].
Neu5Ac and KDN (derived from the condensation of Man-6-P, as an
alternative to ManNAc-6-P for Neu5Ac) are believed to be the precursors
for all other members in the sialic acid family. The different enzymatic
modifications of Neu5A and KDN to form the other members of the family
take place in different cellular compartments.
3.2 Sialyltransferases
The sialyltransferases represent a group of glycosyltransferases that add
sialic acid residues to an acceptor glycoconjugate terminating in Gal, Gal-
NAc, or another sialic acid. The different mammalian sialyltransferases
identified are membrane-bound enzymes present in the Golgi apparatus,
and together with the other glycosyltransferase residents of the Golgi
apparatus, they share a type II architecture. All mammalian sialyl-
transferases characterised today are composed of a short N-terminal cyto-
plasmic domain, a transmembrane domain, a stem region, and a catalytic
domain [74]. The length and amino acid composition of the catalytic
domains of the different sialyltransferases are well conserved, which greatly
assists in the identification of new sialyltransferases. In contrast, the stem
region of the protein is less conserved and the length can vary between
proteins, which is believed to influence the substrate specificity of the
acceptor sugar [74, 75].
Members of the sialyltransferase family are grouped into subfamilies on
the basis of the introduced linkage type (a2,3, a2,6, a2,8) between the donor
sialic acid and the accepting sugar. Since 1996, a universal nomenclature is
Carbohydr. Chem., 2012, 38, 75–93 | 85

being used, which encompasses four components. It discloses the protein
family, linkage type, the acceptor sugar, and the identification number of
the protein within the subgroup. As an example; the b-galactoside a2,3-
sialyltransferase-three is called ST3Gal-III, where ST stands for sialyl-
transferase, 3 represents a2,3 linkage, Gal is the acceptor sugar galactose,
and III indicates that it has been the third gene identified within this sub-
family [76].
3.2.1 a2,3 sialyltransferase sub-family. The most abundant linkage
found in mammalian oligosaccharides is the a2,3 linkage, and to date, six
a2,3 sialyltransferases have been identified in mice and humans. The six
members of the family can be subdivided into two different groups based on
substrate specificity and sequence similarity. The first group consists of the
ST3Gal transferases-I and -II that prefer to use core 1 O-glycans
(Galb1,3GalNAc) on glycoproteins or glycolipids as an acceptor substrate
[74]. ST3Gal-I is expressed in many human tissues, including the heart,
placenta, lungs, and peripheral blood lymphocytes [77]. It is responsible for
the sialylation of the core 1 O-glycans on T-cells. St3gal1–deficient mice
showed increased apoptosis of CD8 þ T-cells, resulting in a nearly complete
deficiency of this cell type [78]. ST3Gal-II is among others expressed in the
heart, liver, and lymphoid tissues of humans, and has a substrate preference
for glycolipids over glycoproteins in vitro [79].
The second group comprises the ST3Gal transferases-III, -IV, -V, and
-VI. ST3Gal-III and -IV can both utilise Galb1,3GlcNAc and Galb1,4Glc-
NAc as acceptor glycans, but their substrate preference differs. ST3Gal-III
prefers the type I lactosamine (Galb1,3GlcNAc) glycan and is expressed in
the liver, testis and fetal tissues. ST3Gal-IV prefers type 2 oligosaccharides
(Galb1,4GlcNAc), which can be further processed to the tetrasaccharide
sialyl Lewis X (sLeX), as acceptor glycans and is strongly expressed in the
placenta, testis, and ovary [74, 77]. Selectin binding studies in St3gal3- and
St3gal4-deficient mice have shown that ST3Gal-IV is essential for the bio-
synthesis of active selectin ligands, whereas ST3Gal-III is not. St3gal4-
deficient neutrophils exhibited a reduction of E- and P-selectin-mediated
rolling in vitro when compared to wild-type neutrophils, while no difference
to wild type was found for rolling of St3gal3-deficient neutrophils [80].
ST3Gal-V is expressed in the testis, brain, and skeletal muscles and utilises
lactosylceramide (Galb1,4Glc-Cer) as a substrate and has not been reported
to contribute to leukocyte recruitment [75, 77]. In contrast, ST3Gal-VI uses
type 2 lactosamines (Galb1,4GlcNAc) on glycoproteins and glycolipids as
an acceptor substrate and is involved in the synthesis of sLex. However, the
contribution of this enzyme to leukocyte rolling and adhesion is currently
unclear.
3.2.2 a2,6 sialyltransferase sub-family. a2,6 linkages are less abundant
in nature than a2,3 linkages, and eight mammalian a2,6 sialyltransferases
have been identified. Based on substrate specificity, two groups of a2,6
sialyltransferases can be distinguished, the ST6Gal group and the
ST6GalNAc group. ST6Gals transfer sialic acids from CMP-sialic acid to
the acceptor sugar galactose of glycans. Two enzymes belonging to this
group have been discovered in mouse and human, ST6Gal-I and ST6Gal-II.
86 | Carbohydr. Chem., 2012, 38, 75–93

Both enzymes typically utilise the Galb1,4GlcNAc structure on carbohy-
drate decorations of proteins as acceptor substrate, but ST6Gal-I has a
broader substrate specificity [74]. The rise in expression of ST6Gal-I during
inflammation is believed to be a vital part of the systemic inflammatory
response [81].
The second group comprises the ST6GalNAc transferases I-VI that utilise
N-acetylgalactosamine (GalNAc) as an acceptor sugar. They can add sialic
acids to Galb1,3GlcNAc and Galb1,4GlcNAc motifs on glycans, but their
substrate preference varies [74]. To our knowledge, none of the ST6Gal-
NAcTs has been reported to significantly modulate innate or adaptive
immune responses, and therefore ST6GalNacTs will not be further
discussed.
3.2.3 a2,8 sialyltransferase sub-family. All six identified a2,8-sialyl-

transferases in mice and in humans link sialic acids to terminal acceptor
sialic acid residues on glycoconjugates. The Siaa2,8Sia sequence is com-
monly found in various gangliosides, which are sialic acid containing
sphingolipids that can play an important role in immunology. Sporadically,
the Siaa2,8Sia sequence is seen in carbohydrate decorations of proteins as
well [82]. Like the ST3 and ST6 subfamilies, the ST8 subfamily can be
segmented into two groups based on substrate specificity and sequence
similarity. The first group contains three enzymes, ST8Sia-I, -V, and -VI.
ST8Sia-I, better known as GD3 synthase, and ST8Sia-V are both
involved in the biosynthesis of gangliosides, but show opposite substrate
specificities. ST8Sia-I displays strong activity towards GM3, an
intracellular regulator of apoptosis in T lymphocytes, but hardly any
activity towards GD1a (co-receptor in Toll-like receptor 2 signalling
[83]), GT1b, and GD3. ST8Sia-V on the other hand, uses GD1a, GT1b,
and GD3 as acceptor substrates, but shows low activity towards GM3
[82]. In contrast to these two enzymes, ST8Sia-VI utilises O-glycans of
glycoproteins as acceptor substrates and shows low activity towards gang-
liosides [74].
The second group consists of three members, ST8Sia-II, -III, and -IV.
ST8Sia-III uses the Siaa2,3Galb1,4GlcNAc glycan chain on N-glycans
of glycoproteins and glycolipids as the sialic acid acceptor. It is also
believed to have the ability to biosynthesise Siaa2,8Siaa2,3Gal on certain
O-glycans [74]. ST8Sia-II and -IV are polysialic acid synthases that catalyse
polysialic acid formation on N-glycans of neural cell adhesion molecules
(NCAMs) [74].
3.2.4 Sialidases. Sialidases, which are also termed neuraminidases

(although this name is now generally reserved to denote sialidases from
viruses [46]) are very common in nature. They can be found in many viral-,
bacterial-, invertebrate-, and mammalian species. Sialidases catalyse the
hydrolytic cleavage of sialic acids from glycoconjugates. The different
enzymes within the sialidases family show different biochemical properties,
such as substrate preference, kinetics, and binding affinities. Recently,
membrane-bound sialidases on neutrophils have been described to regulate
leukocyte adhesion behaviour during inflammation [84].
Carbohydr. Chem., 2012, 38, 75–93 | 87

4 Sialylation-dependent functions of signalling and adhesion relevant
molecules
4.1 Selectin ligands
Selectins were the second group of sialic-acid binding proteins discovered.
Selectin - selectin ligand interactions are essential for leukocyte recruitment
to sites of inflammation. Synthesis of functional selectin ligands is
dependent on various glycosyltransferases which were discussed in Section 2
and 3, and are summarised in Table 1. Deficiency of one of these glyco-
syltransferases leads to more or less pronounced defects in selectin-mediated
rolling which in turn can severely affect leukocyte recruitment during
inflammation and homing of lymphocytes into secondary lymphatic tissue
[18, 50, 57, 58].
To explore the exact role of sialic acids in selectin-selectin ligand inter-
actions, mice deficient in several sialyltransferases have been generated and
their phenotypes studied. Since sialic acids found in the sLeX motif of
selectin ligands are a-2,3 linked to the underlying galactose residue, the
primary focus has been on studying the role of a-2,3 sialyltransferases on
leukocyte recruitment [18, 85]. Three sialyltransferases have been discussed
to contribute to the generation of functional selectin ligands: ST3Gal-III,
ST3Ga-IV, and ST3Gal-VI. Neutrophils of St3gal3-deficient mice showed
no difference in binding to either P-selectin or E-selectin suggesting that
ST3Gal-III is not relevant in the synthesis of functional selectin ligands on
neutrophils [61]. On the contrary, neutrophils of St3gal4-deficient mice
showed a significant reduction in binding to all three selectins in vitro [80].
However, as pre-treatment of St3gal4-neutrophils with sialidase further
reduced binding of the cells to the different selectins, other sialyltransferases
are likely to be involved in the generation of functional selectin ligands. This
is supported by additional intravital microscopic experiments in inflamed
cremaster muscle venules of St3gal4-deficient mice where a much milder E-
selectin dependent rolling defect than under in vitro conditions was observed
with a moderately reduced number of rolling leukocytes on E-selectin and
an increase in E-selectin-dependent rolling velocities [80]. P-selectin
dependent rolling in vivo was not affected at all in the absence of ST3Gal-IV
[80]. These results suggest that for E- and P-selectin ligand activity, other
sialyltransferases might compensate for the loss of ST3Gal-IV [61]. Besides
E- and P-selectin, L-selectin also contributes to leukocyte rolling and
leukocyte recruitment during inflammation in vivo [86, 87]. Using St3gal4-
deficient mice, it has been shown that L-selectin-mediated rolling of neu-
trophils on inflamed cremaster muscle venules was almost completely absent
[62]. Since L-selectin-mediated rolling of neutrophils requires leukocyte-
expressed PSGL-1, these findings suggest that PSGL-1 binding activity
towards L-selectin is strongly dependent on ST3Gal-IV [62].
As described in Section 3, the sialyltransferase ST3Gal-I uses core 1-
decorated O-glycans (Galb1,3GalNAc) as acceptor substrate and therefore
is not likely to be involved in the synthesis of the sLeX motif on core 2-
carrying O-glycans [74]. But, interestingly, St3gal1-deficient mice showed
that the lack of ST3Gal-I activity is accompanied by a strong up-regulation
of sLeX-containing core 2-branched O-glycans, leading to better binding of
88 | Carbohydr. Chem., 2012, 38, 75–93

the selectins to its selectin ligands which indicates that ST3Gal-I is
competing with the glucosaminyltransferase core 2-GlcNAcT-1 for the
same acceptor substrate [78].
The sialyltransferase ST3Gal-VI has been connected to contribute to the
biosynthesis of sLeX. However, ST3Gal-VI is not highly expressed in
hematopoietic cells, including leukocytes [88]. The exact role of ST3Gal-VI
in selectin - selectin ligand interactions in vivo is currently unknown, but
St3gal6-deficient mice have been generated recently. Investigations in those
mice will soon elucidate how ST3Gal-VI contributes to selectin ligand
activity in vivo [46].
4.2 Sialylation and chemokine receptors

Posttranslational glycosylation is not only vital for selectin ligand activity,
but also plays a major role in the proper function of chemokine receptors
[18, 41]. About a decade ago, the first studies identifying a role of post-
translational sialylation on chemokine receptor binding to chemokines have
been published [89]. In these first studies, a canine thymocyte cell-line was
used that was transfected with either wild-type chemokine receptor 5
(CCR5) or one of the various CCR5 mutants in which specific serine or
threonine residues had been exchanged with alanine to remove putative
O-glycosylation sites [89]. It was shown that binding of CCR5 to chemokine
was strongly dependent on a sialic acid-carrying O-glycan, which was linked
to serine 6 at the N-terminus of CCR5 [89].
Subsequently, the important role of posttranslational glycosylation in
chemokine-triggered firm leukocyte arrest in vivo was published [41]. Using
intravital microscopy, the function of ST3Gal-IV on CXCR2-mediated
leukocyte adhesion was investigated. These studies revealed that leukocyte
arrest in cremaster muscle venules of wild-type mice was rapidly and
strongly induced after systemic injection of CXCR2-binding chemokines
(CXCL1 or CXCL8), whereas in St3gal4-deficient mice leukocyte adherence
to the vessel wall was dramatically reduced following systemic injection of
CXCL1 or CXCL8. Additional studies showed that neutrophils isolated
from St3gal4-deficient mice exhibit reduced binding to the chemokines
CXCL1 and CXCL8 [41]. These studies strongly suggest that ST3Gal-IV-
mediated sialylation is critical for CXCR2-induced firm neutrophil adhe-
sion during inflammation [18,41]. Further studies are warranted aiming to
explore the potential role of ST3Gal-IV in regulating the function of other
chemokine receptors in neutrophils and other leukocyte subpopulations
(monocytes, dendritic cells, eosinophils, and lymphocytes) [85].
4.3 Integrins and their ligands

The importance of integrins in the leukocyte recruitment cascade to
inflammatory sites is illustrated by several genetic disorders in humans, in
which defects in integrin expression and/or functions occur. Genetic defects
in adhesion are collectively called leukocyte adhesion deficiencies (LADs).
The most common form of LAD, called LAD-I, is characterised by
mutations in the ITGB2 gene, encoding the leukocyte-expressed b2 integrin
subunit. The different mutations can lead to loss or reduced presence
of b2 integrins on the cell surface and defects in b2 integrin signalling.
Carbohydr. Chem., 2012, 38, 75–93 | 89

LAD-I patients suffer from impaired wound healing and recurrent bacterial
infection [11, 60].
Although posttranslational glycosylation of integrins has been known for
quite some time, its functional role for leukocyte adhesion only begins to
emerge [90]. Woodard-Grice recently showed that the sialyltransferase
ST6Gal-I negatively regulates the activity of the integrin VLA-4 on
monocytes by sialylation of the b1 integrin’s extracellular domain. Inter-
estingly, the activity of ST6Gal-I is controlled by the b-site APP-cleaving
enzyme 1 (BACE1) secretase, which down-regulates ST6Gal-I through
cleavage and shedding of ST6Gal-I from the cell surface. In another report,
loss of a2,6-sialylation was demonstrated to cause significant changes in the
conformation of b1 integrin with reduced accessibility of the ligand binding
region suggesting an important regulatory function of a2,6 sialylation on
integrin function [91].
4.4 Siglecs
At least three classes of lectins involved in leukocyte recruitment have been
described in the literature: C-type lectins (selectins), siglecs (I-type lectins),
and galectins [92, 93]. Two of them, selectins and siglecs rely on sialic acid
containing ligands. The role sialic acids play on proper binding of selectins
to its ligands has been described above. In this paragraph the role of siglecs
and their ligands will be discussed in the context of leukocyte recruitment.
Siglecs (Sia-recognising immunoglobulin superfamily lectins) belong to
the I-type lectin family as they fulfil the two requirements necessary for
I-type lectins: 1) they are members of the immunoglobulin-superfamily of
proteins and 2) they recognise terminal sugars (sialic acids in case of siglecs)
on glycoconjugates. Siglecs are type I transmembrane proteins expressed by
all cell types of the innate and adaptive immune system [94–96]. Depending
on the presence or absence of a cytoplasmic signalling domain, they can be
involved in cell-cell interactions and/or intercellular communication [97]. In
addition, siglecs are used as receptors by some pathogenic microorganisms
in a process known as molecular mimicry, where the bacterium decorates its
glycolipids on the cell surface with Neu5Ac to promote interactions with
siglecs that function as immune receptors [95].
Structurally, siglecs contain a transmembrane domain, a variable number
of extracellular C2-set immunoglobulin domains, and an N-terminal V-
set-immunoglobulin domain that contains a conserved arginine residue for
sialic acid binding. On the basis of sequence similarities and evolutionary
conservation, siglecs can be divided into two groups. The first group com-
prises the CD33-related siglecs (CD33/siglec3, siglec-5 till -11, siglec-14 in
humans, CD33/siglec-3, siglec-E, -F, -G, and -H in mice) and the second
group consists of siglec-1 (sialoadhesin or Sn), siglec-2 (CD22), MAG
(siglec-4), and siglec-15. Distinctive siglecs are expressed by different cell-
lines of the hematopoietic or immune system. For example, siglec-1 can be
found on macrophages and to a lesser degree on granulocytes [98], CD33
(siglec-3) is expressed on neutrophils, monocytes and macrophages [99].
Siglecs from both groups recognise sialic acid residues on N- or O-linked
carbohydrate decorations of proteins or on glycolipids, distinguishing also
the sialic acid-linkage type they bind to (a2,3, a2,6, or a2,8). Due to the high
90 | Carbohydr. Chem., 2012, 38, 75–93

level of sialic acids on cell surfaces, siglecs are mostly bound to sialic acid
residues in cis, which leads to a masking effect of the sialic acid binding site
on siglecs under homeostatic conditions. This is true for most siglecs. An
exception is siglec-1 which has 17 Ig-like domains and extends well beyond
the cell surface (roughly 50 nm) [100]. So it can be assumed that the sialic
acid binding site of siglec-1 is not masked and might be able to interact with
sialic acid containing ligands on other cells (trans-interaction). As siglec-1
does not contain any cytoplasmic signalling domain, it may therefore serve
as a mere adhesion molecule. This is indirectly supported in a report by
Jiang et al. [101]. Using siglec-1-deficient mice, they found an attenuated
initial disease progression in an autoimmune uveoretinits model suggesting
a pro-inflammatory function of siglec-1 in this T-cell driven inflammation
model [101]. Although it is still unclear, how siglec-1 exerts this pro-
inflammatory function, there is evidence that siglec-1 binds to glycoproteins
(including PSGL-1) containing terminal a2,3-linked sialic acids, which
might be an indication that siglec-1 could participate in leukocyte recruit-
ment into inflamed tissue. Future studies including in vivo imaging will be
needed to further elucidate the exact molecular mechanisms by which siglec-
1 regulates immune cell recruitment.
5 Conclusion and outlook

In the last few years, posttranslational sialylation has emerged as an
important modification of adhesion molecules and adhesion relevant
molecules impacting leukocyte recruitment into tissue during an inflam-
matory response. While a2,3 sialylation has been well-established as crucial
modulator of selectin ligand function and hence leukocyte capture and
rolling, recent findings also reveal a role of a2,3 sialylation on chemokine
receptor function and chemokine-triggered firm leukocyte arrest. Further-
more, several investigations provide now strong evidence that a2,6 sialyla-
tion of b1 integrin negatively regulates b1 integrin activation stretching the
contribution of posttranslational glycosylation from the initial rolling step
to firm leukocyte arrest and postarrest modifications. These new findings
emphasize the importance of carbohydrate decoration on the regulation of
an inflammatory response and also offer new interesting therapeutic
approaches in the treatment of acute and chronic inflammatory diseases.
References
1 H. Dutrochet, Wilhelm Engelmann Verlag, 1824.
2 R. Wagner, Leopold Voss Verlag, 1839.
3 E. C. Butcher, Cell, 1991, 67, 1033.
4 K. Ley, et al., Nat Rev Immunol, 2007, 7, 678.
5 T. A. Springer, Cell, 1994, 76, 301.
6 G. S. Kansas, Blood, 1996, 88, 3259.
7 D. Vestweber and J. E. Blanks, Physiol Rev, 1999, 79, 181.
8 M. Sperandio, FEBS J, 2006, 273, 4377.
9 A. Zarbock and K. Ley, Am J Pathol, 2008, 172, 1.
10 A. Zarbock, et al., Blood, 2011, 118, 6743.
11 C. L. Abram and C. A. Lowell, Annu Rev Immunol, 2009, 27, 339.
12 M. Phillipson, et al., J Exp Med, 2006, 203, 2569.
Carbohydr. Chem., 2012, 38, 75–93 | 91

13 L. Stephens, et al., Curr Biol, 2008, 18, R485.
14 M. Sperandio, et al., Methods Enzymol, 2006, 416, 346.
15 Z. Liu, et al., J Exp Med, 2010, 207, 2975.
16 D. Vestweber, Immunol Rev, 2007, 218, 178.
17 R. D. Cummings and R. P. McEver, C-type lectins, In: Essentials of Glyco-
biology, 2nd edition, 2009.
18 M. Sperandio, et al., Immunol Rev, 2009, 230, 97.
19 L. Xia, et al., J Clin Invest, 2002, 109, 939.
20 M. Sperandio, et al., J Exp Med, 2003, 197, 1355.
21 A. Hidalgo and P. S. Frenette, Immunity, 2007, 26, 753.
22 M. Matsumoto, et al., J Immunol, 2005, 175, 8042.
23 P. Sundd, et al., Biorheology, 2011, 48, 1.
24 E. B. Finger, et al., Nature, 1996, 379, 266.
25 B. T. Marshall, et al., Nature, 2003, 423, 190.
26 T. S. Olson and K. Ley, Am J Physiol Regul Integr Comp Physiol, 2002, 283,
R7.
27 A. Viola, and A. D. Luster, Annu Rev Pharmacol Toxicol, 2008, 48, 171.
28 N. Hogg, et al., Nat Rev Immunol, 2011, 11, 416.
29 F. Ye, et al., Blood, 2012, 119, 26.
30 R. Evans, et al., J Cell Sci, 2009, 122, 215.
31 M. T. Huang, et al., Blood, 2006, 107, 4721.
32 D. Frommhold, et al., Blood, 2010, 116, 841.
33 S. Nourshargh, et al., Nat Rev Mol Cell Biol, 2010, 11, 366.
34 W. A. Muller, Annu Rev Pathol, 2011, 6, 323.
35 J. D. Marth, Nat Cell Biol, 2008, 10, 1015.
36 R. A. Dwek, Chem Rev, 1996, 96, 683.
37 K. Marino, et al., Nat Chem Biol, 2010, 6, 713.
38 H. H. Freeze, Nat Rev Genet, 2006, 7, 537.
39 A. Varki and J. B. Lowe, Biological Roles of Glycans, In: Essentials of Gly-
cobiology, 2nd edition, 2009.
40 J. D. Marth and P. K. Grewal, Nat Rev Immunol, 2008, 8, 874.
41 D. Frommhold, et al., J Exp Med, 2008, 205, 1435.
42 A. V. Woodard-Grice, et al., J Biol Chem, 2008, 283, 26364.
43 G. Reuter and H. J. Gabius, Cell Mol Life Sci, 1999, 55, 368.
44 K. Uchimura and S. D. Rosen, Trends Immunol, 2006, 27, 559.
45 K. Ohtsubo and J. D. Marth, Cell, 2006, 126, 855.
46 A. Varki and R. Schauer, Sialic Acids, In: Essentials of Glycobiology, 2nd
edition, 2009.
47 L. Tu and D. K. Banfield, Cell Mol Life Sci, 2010, 67, 29.
48 M. Sperandio, et al., J Immunol, 2001, 167, 2268.
49 M. Sperandio, et al., Blood, 2001, 97, 3812.
50 L. G. Ellies, et al., Immunity, 1998, 9, 881.
51 M. Asano, et al., Blood, 2003, 102, 1678.
52 B. Hansske, et al., J Clin Invest, 2002, 109, 725.
53 M. Guillard, et al., J Pediatr, 2011, 159, 1041.
54 B. Ma, et al., Glycobiology, 2006, 16, 158R.
55 D. J. Becker and J. B. Lowe, Glycobiology, 2003, 13, 41R.
56 C. M’Rini, et al., J Exp Med, 2003, 198, 1301.
57 J. W. Homeister, et al., Immunity, 2001, 15, 115.
58 P. Maly, et al., Cell, 1996, 86, 643.
59 W. Weninger, et al., Immunity, 2000, 12, 665.
60 S. Hanna and A. Etzioni, Ann N Y Acad Sci, 2012,
61 L. G. Ellies, et al., Proc Natl Acad Sci USA, 2002, 99, 10042.
92 | Carbohydr. Chem., 2012, 38, 75–93

62 M. Sperandio, et al., Eur J Immunol, 2006, 36, 3207.
63 A. Varki, Glycobiology, 1992, 2, 25.
64 A. Varki, Nature, 2007, 446, 1023.
65 S. Kelm and R. Schauer, Int Rev Cytol, 1997, 175, 137.
66 R. Schauer, Curr Opin Struct Biol, 2009, 19, 507.
67 T. Angata and A. Varki, Chem Rev, 2002, 102, 439.
68 R. Schauer, Glycoconj J, 2000, 17, 485.
69 O. T. Keppler, et al., Science, 1999, 284, 1372.
70 S. Hinderlich, et al., J Biol Chem, 1997, 272, 24313.
71 S. M. Lawrence, et al., J Biol Chem, 2000, 275, 17869.
72 P. Maliekal, et al., Glycobiology, 2006, 16, 165.
73 M. Eckhardt, et al., Proc Natl Acad Sci USA, 1996, 93, 7572.
74 S. Takashima, Biosci Biotechnol Biochem, 2008, 72, 1155.
75 A. Harduin-Lepers, et al., Biochimie, 2001, 83, 727.
76 S. Tsuji, et al., Glycobiology, 1996, 6, v.
77 H. Kitagawa and J. C. Paulson, J Biol Chem, 1994, 269, 1394.
78 J. J. Priatel, et al., Immunity, 2000, 12, 273.
79 V. Giordanengo, et al., Eur J Biochem, 1997, 247, 558.
80 L. G. Ellies, et al., Blood, 2002, 100, 3618.
81 J. C. Jamieson, et al., Comp Biochem Physiol B, 1993, 105, 29.
82 S. Tsuji, J Biochem, 1996, 120, 1.
83 S. Liang, et al., J Biol Chem, 2007, 282, 7532.
84 C. Feng, et al., J. Leukoc. Biol., 2011, 90, 313.
85 M. Sperandio, Ann N Y Acad Sci, 2012,
86 M. L. Arbones, et al., Immunity, 1994, 1, 247.
87 K. Ley, et al., Journal of Experimental Medicine, 1995, 181, 669.
88 T. Okajima, et al., J Biol Chem, 1999, 274, 11479.
89 N. Bannert, J Exp. Med., 2001/12/3, 194, 1661.
90 M. E. Janik, et al., Biochimica Biophysica Acta, 2010, 1800, 545.
91 Y. Liu, et al., Proteins, 2008, 73, 989.
92 G. R. Vasta, Nat Rev Microbiol, 2009, 7, 424.
93 T. K. Dam and C. F. Brewer, Glycobiology, 2010, 20, 270.
94 L. G. Baum and P. R. Crocker, Immunol Rev, 2009, 230, 5.
95 P. R. Crocker, et al., Nat Rev Immunol, 2007, 7, 255.
96 A. Varki and T. Angata, Glycobiology, 2006, 16, 1R.
97 L. Nitschke, Immunol Rev, 2009, 230, 128.
98 P. R. Crocker, et al., EMBO J, 1994, 13, 4490.
99 D. Simmons, et al., Nature, 1988, 331, 624.
100 A. Varki and P. R. Crocker, I-type lectins, In: Essentials of Glycobiology, 2nd
edition, 2009.
101 H. R. Jiang, et al., J Immunol, 2006, 177, 2258.
Carbohydr. Chem., 2012, 38, 75–93 | 93

Glycoengineering of protein-based
therapeutics
Sandrine Donadio-Andréi,w Chloé Iss,w Nassima El Maı̈,
Valérie Calabro and Catherine Ronin*
DOI: 10.1039/9781849734769-00094
Introduction
N-linked glycosylation is widely considered as the most complex
modification of proteins in the eukaryotic kingdom [1]. N-glycans play a
crucial role in the biosynthesis, folding, stability, biological activity and
metabolic clearance of the proteins to which they are attached [2, 3]. They
control multiple key biological processes including cell adhesion, receptor
activation, signal transduction and endocytosis [3–5]. In addition, they are
highly versatile structures that have been shown to be associated with in a
wide panel of pathological events, especially in infection, cancer, immuno-
logical diseases, metabolic and neurodegenerative disorders [6–9].
The biosynthetic machinery which is in charge of assembling oligo-
saccharides onto proteins requires a large repertoire of enzymes [10].
Expression of glycosylation enzymes appear to vary among organisms,
from cell type to cell type within a single organism as well as with cellular
environment and stage. The composition and the linkage(s) of mono-
saccharide residues in the oligosaccharide chains are thus related to the
species. As a result, glycoproteins expressed in animal cells could be
therefore strongly immunogenic if administered in humans missing these
structures [11]. The nature of the oligosaccharide chain profoundly affects
stability and pharmacokinetic properties of most glycosylated therapeutic
proteins [3, 12, 13]. N-glycosylation is therefore a major concern for the
development of biopharmaceutical proteins.
At present, over 70% of the marked protein-based therapeutics are
glycosylated proteins and several hundreds of these products are being
developed. They consist in monoclonal antibodies (mAb) and mAbs-based
products, proteins for replacement therapies (hormones, growth factors,
enzymes) and cytokines. It is important for protein-based therapeutics to be
produced with a glycosylation optimal for their function and similar to their
human counterparts to minimize risk of immunogenicity that may com-
promise treatment efficacy. Another issue for the production of bio-
pharmaceutical glycoproteins is the glycoform profile [14]. The inherent
properties of the biosynthetic machinery leads to multiple N-glycoprotein
variants for which the glycan pattern can undergo substantial variation with
changes in cell culture conditions. A major challenge for biopharmaceutical
industry is to control this extensive heterogeneity and obtain batch-to-batch
consistency for optimum bioefficacy and reliable safety evaluation of their
products [15, 16].
Siamed’Xpress, Aix-Marseille Univ., 3 Pl.V.Hugo, 13331-Marseille, France.

E-mail: catherine.ronin@siamedxpress.com
w
equal contribution.
94 | Carbohydr. Chem., 2012, 38, 94–123

Protein may display variable glycans and therefore, their exposed surfaces
differ among glycoforms depending on the oligosaccharide structure. This
may affect antibody recognition [17] and may be an important issue in
assaying or tracing a drug candidate. It is highly desirable to generate the
right glycoform (when generating either antibodies against a human
glycosylated target for diagnostics or for developing therapeutic anti-
bodies). In diagnostics, the desired glycoform(s) of a protein marker should
presumably be identical to the circulatory form(s) of the analyte to ensure
epitope similarity. Therapeutics antibodies must be selected against the
appropriate target glycovariant(s) to display optimal pharmacological
activity. Additionally, N-glycosylation changes are associated with diseases
such as cancer and inflammation and measuring diseased glycoform(s)
could be of major diagnostic and prognostic significance [18, 19].
Currently, therapeutic glycoproteins are predominantly produced in
rodent expression systems such as Chinese Hamster Ovary (CHO), mouse
NS0 or mouse SP2/0 cell lines because their glycosylation machinery is the
closest to that of human cells [20, 21]. However, these systems are not
perfect, some sugar components are missing and instead, some rodent-type
residues are added and this significantly increases the likelihood of immuno-
genic reactions. Efforts are currently being made to generate biotherapeutics
with ‘‘humanized’’ glycosylation to reduce immunogenicity, optimize
bioactivity and generate a consistent glycoform profile with limited het-
erogeneity [22, 23]. Here, we review the issues associated with protein gly-
cosylation for therapeutic and diagnostic applications and the main
approaches to date to address these issues.
1 Impact of N-glycosylation on therapeutic proteins

1.1 N-glycosylation achieved by heterologous expression systems
N-glycosylation is not directly encoded by the genome. It is a mechanism
conserved across eukaryotes and represents the most widely distributed
glycan addition in proteins [1]. It requires the tripeptide consensus site Asn-
X-Ser/Thr where X could be any amino acid except proline; the oligo-
saccharide chain is attached to the asparagine residue. The synthesis occurs
co-translationally in the lumen of the endoplasmic reticulum and post-
translationally in the Golgi apparatus. It appears tightly regulated and the
rules that control this highly complex process are not yet fully understood.
First, the glycoprotein sequence and the structure surrounding the consensus
sequence is thought to determine glycan addition. Occupied N-glycosylation
sites are composed of a common core pentasaccharide constituted by two
GlcNAc and three mannose residues and then, depending on the organism,
the cells and/or the substrate availability, different sugars are being added to
form bi-, tri- or tetraantennary glycans with various terminal residues as
shown in Fig. 1. A series of glycosyltransferases and glycosidases are
involved in the process. They act sequentially to build up the oligosaccharide
chain and compete with each other for both the available protein acceptors
and nucleotide donors. Their expression is regulated in function of cell type,
development stage, cell environment and culture conditions and their com-
bined actions deliver a wide repertoire of glycoforms within each cell type.
Carbohydr. Chem., 2012, 38, 94–123 | 95

Fig. 1 Typical structure of the common mannose-rich precursor of all human N-glycans and
the three classes of N-linked glycans. Dashed bars depict the elongation of the pentasaccharide
core with complex antennae, generating a wide panel of N-glycans. Abbrevations: GnT: N-
acetylglucosaminyltransferase, FUT8: a1,6 fucosyltransferase, ST: sialyltransferase, GalT:
galactosyltransferase.
Several expression systems are available for production of therapeutic

proteins: bacteria, yeasts, plants, insect cells, mammalian cells and trans-
genic animals. So far, mammalian cells remain the system of choice for
biotherapeutic glycoprotein production because of its reliability [19]. The
use of other systems as an alternative to mammalian cells has been mainly
hindered by their limited capacity to produce glycoprotein with human-like
96 | Carbohydr. Chem., 2012, 38, 94–123

glycosylation [21]. Plant and insect glycans contain b1,2-xylose and core
a1,3-fucose which elicit strong IgG and IgE responses when administered to
humans [24, 25]. In addition, plant cells poorly transfer outer galactose
(Gal) and do not add sialic acid, leaving essentially pauci-mannosidic or
hybrid glycans which are rapidly cleared by a serum mannan-binding lectin
(MBL) following injection. Yeast cells have not been intensively used
for protein drug production because they deliver a strongly immunogenic
O-glycosylation. Yeast cells share however similar early biosynthetic
pathway for N-glycan synthesis with mammals but are unable to carry out
complex glycosylation. Antibodies expressed in yeast show defects in
effector function due to altered glycosylation pattern [13]. These cells have
being intensively engineered with the knockout of genes to eliminate yeast-
specific glycosylation, and introduction of heterologous genes to replicate
the branching pattern and terminal N-glycosylation of human proteins [26].
They are probably the most advanced technology that may compete with
mammalian cells for glycoprotein production.
While much effort has been invested in engineering human-like glycosy-
lation pathways in plants, insect and avian cells [22], production of
therapeutic glycoproteins is still currently performed using mammalian cell
systems and is expected to continue over the next decade. Over the years,
mammalian cells have been manipulated to produce large quantities of
biopharmaceutical proteins and are continuously subject to novel
improvements [27]. The most widely used systems are CHO, NS0 and SP2/0
cell lines in various engineered derivatives. Most approved biopharmaceu-
ticals are produced in CHO and CHO-derived cell lines. A quite attractive
aspect of CHO cells is their proven track record for producing therapeutic
proteins which definitely eases the process of achieving drug approval.
Importantly, a wide knowledge and expertise have been accumulated over
more than a hundred of gene products. Figure 2 illustrates carbohydrate
heterogeneity of such recombinant preparations: branching heterogeneity is
illustrated with human TSH (a) and terminal heterogeneity with human
erythropoeitin (hEPO) (b) Indeed, CHO cells display the wide repertoire of
N-linked glycans but do not select biantennary structure in TSH as in the
native hormone or highly sialylated glycans like in tri-and tetraantennary
glycans in native EPO.
In 2011, the complete sequencing of CHO (K1 variant) genome
was achieved and this will undoubtedly lead to further optimization
for higher productivity and higher product quality [30]. However, CHO
glycosylation machinery differs from that of human and most often pro-
duces a random branching pattern of glycosylation. More generally,
rodent cells usually add terminal Gal a1,3-Gal and provide a low degree of
sialylation which may include N-glycolylneuraminic acid (NeuGc); both
sugars are absent in human glycoproteins. Nonetheless, since their early
approval for the production of recombinant hEPO and human tissue
plasminogen activator (tPA), CHO cells have provided several successful
therapeutic proteins and will remain the industry’s favourite expression
system, at least in the near future. Therefore, they are definitively the
best cell candidates for testing new glycoengineering strategies for drug
optimization.
Carbohydr. Chem., 2012, 38, 94–123 | 97

(a)
100 2792
4588
100
2431
%
Abundance
5037
0
4500 4700 4900 5100 5300 5500
Mass (m/z)
2966
*
2605 3604
3241 3778
2070 3416
* * ** * 4226 4588
0
1000 1800 2600 3400 4200 5000
m/z
(b)
Monosialylated
(31%) Non sialylated
glycoforms
(21%)
Disialylated
Sialylated (19%)
glycoforms Tetrasialylated
(79%) (11%)
Tetrasialylated
(18%)
Fig. 2 (a) MALDI-TOF–MS of N-glycans released from recombinant human TSH glycans
produced in mammalian cells. The composition of the most abundant glycoforms are indicated
above each ion. The oligosaccharides are highly heterogeneouss with biantennary tri-, and
tetra- as well as pentaantennary oligosaccharides with or without inner fucose and terminal
sialic acid. Square represents N-acetylglucosamine (GlcNAc), triangle fucose, circle mannose,
circle galactose, diamond sialic acid to symbolize NeuAc/Gc [28]. (b) Relative amount of sia-
lylated glycans in recombinant human EPO produced in CHO cells [29].
1.2 Impact of glycans on drug bioavailability

The residence time of proteins in blood is determined by various parameters
that characterize each protein and affect its susceptibility to different elim-
ination pathways. These parameters such as protein charge, size, hydro-
phobicity or presence of epitopes mediate protein removal from the
circulation and, therefore, play a major role in therapeutic protein efficacy.
N-glycans may affect all these parameters. Oligosaccharide chains are often
large and bulky and alter the hydrodynamic radius of the protein which can
protect from renal clearance. They also affect the structure, the chemical
and physical stability of the protein carrier. Glycosylation has been shown
to improve protein resistance to pH, chemical and thermal denaturation,
and aggregation [3, 31–33]. There is also ample evidence that oligo-
saccharides protect from proteolysis by covering amino acids and thus
preventing access to proteases [34].
98 | Carbohydr. Chem., 2012, 38, 94–123

Although presence of oligosaccharide can improve protein half-life in serum,
the type of glycans can also promote receptor-mediated metabolic clearance.
The metabolic pathways which recycle or eliminate circulatory glycoproteins
involve lectin receptors, essentially the liver asialoglycoprotein receptor
(ASGP-R) and the mannose receptors i.e. the mannan-binding lectin (MBL)
and the macrophage mannose receptors. Each of them can promote glyco-
protein uptake and degradation within minutes. Mannose receptor can
also bind terminal N-acetylglucosamine (GlcNAc) and cause faster protein
clearance [35]. For this reason, particular attention has been paid to the pre-
sence of high mannose glycoforms during drug production. All heterologous
systems ranging from yeast to mammals, including insect and avian cells pri-
marily synthesize mannose-rich glycosylation as it is the precursor form of the
glycan repertoire. It often happens during fermentation that those mannose
residues are not sufficiently trimmed and induce a liability for drug efficacy.
Asialoglycoproteins with free b-linked Gal residues bind to the ASGP-R
the liver receptor and are cleared from serum with remarkably high capacity
[36]. Capping with terminal sialic acids actually masks Gal residues, pre-
venting binding to the lectin receptor and liver uptake resulting in prolonged
circulatory half-life. Therefore, terminal glycosylation and more especially
protein sialylation is a major challenge during bioproduction because it
affects the pharmacokinetic properties of the drug. Recently, it has been
shown that 6-but not 3-linked sialic acid may also modulate this uptake in
mouse [37]. Indeed, loss of activity due to hepatic clearance has been
observed for several protein drugs, the most striking example being the case
of erythropoeitin [38]. The level of sialylation does not influence circulatory
lifespan solely because of the recognition by the ASGP-R but also because of
it impacts on protein properties. Sialic acids are highly hydrophilic negatively
charged residues that alter the pI and solubility/aggregation of the protein.
While the sialic acid content of recombinant hEPO (rhEPO) has been directly
to the increase in circulatory residence time [39, 40], such property has also
used to design more soluble erythropoietin, Darbepoetin alfa, a hypergly-
cosylated analogue of hEPO that contains two additional N-linked carbo-
hydrates [41]. This product displayed a greater overall negative charge which
extends its circulating half-life by a factor of 3 compare to rhEPO in animal
models.
Sialylation therefore appears to be a key parameter to modulate the
in vivo efficacy of therapeutic glycoprotein by improving pharmacokinetic
profiles. Interestingly, a market surveillance conducted by the Swiss Agency
for Therapeutic Products (Swissmedic) has recently released a study aiming
at comparing the major oligosaccharide profiles of 16 commercially avail-
able therapeutic antibodies [42]. They report that N-carbohydrate structures
were predominantly truncated biantennary glycans, with core fucose. The
extent of galactosylation of these products is limited and they are lacking
sialic acid. Hence, enhancing the sialic acid content in therapeutic protein
preparations has the potential to reduce drastically the dosage and
frequency of administration required in for clinical benefit. The influence of
sialic acid on the aggregation propensity of proteins might be important as
well for formulation and storage as it would prevent aggregation and permit
longer storage.
Carbohydr. Chem., 2012, 38, 94–123 | 99

1.3 Impact of glycans on glycoprotein biopotency
The first therapeutic protein produced in recombinant mammalian cells to
obtain market approval was the recombinant human tissue Plasminogen
Activator. Recombinant tPA is a serine protease which converts plasmi-
nogen into plasmin and induces clot lysis. It is indicated for embolic or
thrombotic stroke. tPA is one prominent example in which biological
properties of a glycoprotein have been shown to depend on the structural
features of its oligosaccharides. Two major glycoforms of recombinant tPA
are produced by CHO cells. One glycoform has three glycosylation sites
occupied whereas the second one has only two [43]. Both tPA bioactivity
and bioavailability vary upon site occupancy. The highly glycosylated
glycoform displays lower clot lysis activity and the lack of one site of gly-
cosylation affects clearance rate [44, 45].
Another well documented example of how glycans modulate protein
structure, stability and function is represented by the immunoglobulin
family, the largest class of therapeutic proteins [13, 46]. Most therapeutic
antibodies are of the IgG1 subtype and contain a single N-linked glycan
attached to Asn297 in each Fc domain. The nature and composition of the
oligosaccharide chains carried by the Fc fragment are known to play a
major role in antibody folding and stability. They maintain the spatial
organization of the CH2 domains in an ‘‘open conformation’’ of the Fc
dimer structure, a conformation promoting interaction with Fc receptors
and binding of the C1q protein of the complement [47, 48]. The biological
mechanisms that are activated by immune complexes depend on IgG
Fc oligosaccharides as well [46]. As in other therapeutic glycoproteins,
N-glycosylation also governs biophysical and pharmacokinetic properties
of antibodies and can induce immunogenic reactions.
In serum-derived human IgGs, most of the oligosaccharides are of the
biantennary complex type with a common pentasaccharide core and
variable additions of fucose, bisecting N-acetylglucosamine (GlcNAc)
and terminal GlcNAc, Gal, and sialic acid. However, alterations of endo-
genous IgG glycosylation have been reported in various immune diseases.
In patients with rheumatoid arthritis (RA), both IgG Fc galactosylation
and sialylation are decreased [49]. IgG Fc glycosylation has been found to
modulate inflammatory activities of such IgGs [50]. Even though the
underlying mechanism is not yet understood and still under debate, this
activity is used in intravenous immunoglobulin (IVIg) treatment for many
autoimmune and systemic inflammatory diseases [51]. Several mechanisms
could be at the origin of the anti-inflammatory activity. Increased level of
asialo, agalacto, core-fucosylated biantennary glycoforms may contribute
to inflammation by activating the lectin pathway. Indeed agalactosylated
glycoforms of aggregated immunoglobulin G induce association with the
MBL and contribute to RA pathology [52]. Likewise, using a mouse model,
Kaneko et al., has demonstrated that sialylated human IgGs have a greater
anti-inflammation activity than non sialylated IgGs [53]. They have
shown that a2,6 sialylated N-glycans of IgG can bind a C-type lectin, SIGN-
R1 or in a mouse model. This interaction initiates an anti-inflammatory
cascade leading to the upregulation of cell surface expression of the inhibitory
Fc receptor, FcyRIIb on inflammatory cells and thereby attenuating
100 | Carbohydr. Chem., 2012, 38, 94–123

autoantibody-initiated inflammation [54, 55]. Recently, it has been
suggested that the a2,6 sialic acid residues can be directly involved in the
induction of apoptosis of B cells via its binding to the sialic acid-binding
receptor CD22 thus contributing to the anti-inflammatory activity of
IVIg [56].
While native glycosylation of human serum IgGs is sufficient for immune
effector functions, marketed recombinant monoclonal antibodies are being
optimized to enhance their therapeutic bioefficacy especially in the field
of cancer or infectious diseases [22, 57]. Indeed, IgG glycoforms display
differences in their ability to trigger effector functions. Various carbohy-
drates like core fucose, bisecting GlcNAc, sialic acid, and mannose residues
affect the binding of IgG to the Fcg receptors and thereby influence
Antibody-Dependent Cellular Cytotoxicity (ADCC) activity [54]. The
absence of core fucose as well as the presence of bisecting GlcNAc have
been found to significantly increase binding of IgG Fc to FcyRIIIa [58–60].
ADCC is thus increased through longer residence time on FcyRIIIa
receptor and out competition with serum IgG. Removal of terminal Gal
reduces Complement-Dependent Cytotoxicity (CDC) by approximately
50% relative to CDC function of unmodified mAbs [61, 62]. IgGs with
mannose-rich glycans are part of the glycoforms of mAbs produced in CHO
cells [63] and show low affinity for the complement C1q protein and low
CDC activity compared to complex type glycoforms [64], indicating that
their presence in an antibody preparation will decrease the efficacy of the
whole product. Alternatively, the presence of terminal sialic acid residues in
the IgG1 Fc oligosaccharides reduces binding affinities to the FcyRIIIa
receptor, thereby decreasing in vivo cytotoxicity [53].
Many new antibody-based therapeutics are currently under development
and it is anticipated that these glycoproteins will represent 30% of the
marketable drugs in the next five years. Thousands of patents have been
issued so far to protect antibody engineering. There is a considerable interest
in manipulating both their polypeptide and oligosaccharide chains to
maximize their anti-infectious, anti-cancer or anti-inflammatory activities.
1.4 Impact of glycans on drug immunogenicity

Another important cause of low bioavailability resides in antigenic
reactions to the drug. Immunogenicity, in addition to pose safety concerns,
affects the efficacy of treatment since antibodies raised by the patients
against a therapeutic protein may mediate neutralization of the drug and its
elimination. Clearance of Infliximab (trade name Remicades), an anti-
TNFa monoclonal antibody, is 2.7 times higher and its half-life 34% weaker
in the presence of anti-Infliximab antibody [65].
Immunogenicity first originates from the absence of glycans [11]. It has
been observed that lack of glycosylation site occupancy may induce poly-
peptide misfolding and create new epitopes. For example, recombinant
human granulocyte-macrophage-colony-stimulating (GM-CSF) factor
has been shown to be more immunogenic when produced in E. coli [66].
Anti-recombinant human GM-CSF antibodies react with peptide regions
which are normally protected by glycosylation in the natural protein.
Indeed, such immunogenicity can lead to serious medical issues as the
Carbohydr. Chem., 2012, 38, 94–123 | 101

protein drugs have a human primary structure. Some patients who were
administered recombinant erythropoietin from CHO cells have developed
antibody against their own erythropoietin. Studies have shown that in such
reactive patients, anti-EPO antibodies were directed against the amino acid
and not the glycans [67, 68]. Carbohydrates added on rhEPO by CHO cells
are different from native EPO and may alter EPO conformation to create
epitopes recognized as foreign by human immune system.
Secondly, non-human glycans elicit an anti-sugar immune response that
also greatly reduces blood circulation time. High NeuGc level is found in
proteins produced in NS0 and SP2/0 cells and to a much lesser extent
in CHO cells. This induces immune responses following administration to
patients and causes rapid removal of therapeutic proteins from the circu-
lation [69–71]. The half-life reduction of therapeutic antibodies in clinical
studies has likewise been associated with the presence of anti-Gal
antibodies. Glycoproteins produced in murine cells mostly carry glycans
terminated in Gal-a1,3-Gal residues, which are absent in proteins produced
by human and CHO cells [72]. Since anti-a1-3-Gal antibodies are abundant
in humans and constitute 1% of circulating antibodies in humans [73],
glycoforms carrying this type of residue are rapidly neutralized and cleared.
Cetuximab (trade name Erbitux), an anti-EGFR mAb approved by the
FDA in 2004 for the treatment of metastatic colorectal cancer, head and
neck cancers, has been produced in SP2/0 cell line. 22% of patients treated
with Cetuximab displayed severe hypersensitivity reactions. Immunological
testing revealed that patient antibodies against Cetuximab were specific for
Gal-a1,3-Gal present on the Fab portion of the therapeutic antibody [74].
Alternatively, terminal sialic acid has been shown to prevent recognition by
T- and B-cell receptors and offer an opportunity to generate anti-drug
neutralizing antibodies [75].
Even though in most instances, the clinical benefit derived from therapies
using biologicals outweighs the risk of developing immunogenic reactions,
immune recognition of protein drugs is increasingly becoming a key issue
for effective long-term treatment. It should be further emphasized that this
concern is more particularly relevant for glycoproteins indicated for chronic
inflammatory diseases such as RA because such patients already react to
their own immunoglobulins [22].
1.5 Impact of glycans on protein folding and epitope expression

Exposure of peptidic epitopes at the surface of a protein may be modulated
by glycosylation in different ways. Based on their volume, branching and
polarity, glycans can alter protein conformation. Glycoprotein drugs may
fail in binding their target in situ if their glycosylation state has altered their
biological activity. Similarly, changes in glycosylation of a diseased target
may significantly diminished recognition by a therapeutic antibody that has
been designed toward a recombinant glycoform pattern which fails to
represent the native target the natural target. In this respect, lack of epitope
similarity between natural and recombinant or even among recombinant
preparations of a given glycoprotein antigen has been observed [76]. It is
therefore worth producing glycoproteins with a defined glycoprofile
adapted to their biomedical application.
102 | Carbohydr. Chem., 2012, 38, 94–123

However, it has been observed over the various generations of diagnostic
tests that routine immunoassays of glycosylated analytes were discordant
among manufacturers [77]. Regulatory agencies have been sensitized to
current discordances and in the context of cost reduction, now request that
assays should be harmonized and standardized to improve accuracy and
reproducibility of biological tests. About a hundred of glycoprotein markers
are routinely measured in laboratory medicine and there is a clear need for
early diagnostic and/or decision-making of personalized drug treatment in
many diseases. The International Federation of Clinical Chemistry has
elected the measurement of Thyroid Stimulating Hormone (TSH) to initiate
this challenging task as this protein-based marker is the 1st protein marker
prescribed worldwide. Over the 25 past years, many studies have reported
inter-and intra-immunoprocedure discordances in TSH measurements
[78–87]. Recent studies from our group revealed that these discrepancies can
be attributed to TSH glycosylation. The pituitary hormone exists as a mix-
ture of a dozen of GalNAc-sulfated glycoforms which substantially differ in
their biological and immunological properties [88]. In hypothyroid patients,
serum TSH is highly enriched in sialic acids as hypothyroidism develops and
immunoassays lack accuracy as most antibodies bind the circulating analyte
better than the extractive antigen [76, 89, 90]. Essentially, fucose and sialic
acid are modulating the expression of epitopes specific for blood TSH: a 2-3
fold increase in antibody binding was observed when a2,6-linkage sialic acid
or when fucose is lacking. These findings led us to conclude that glycosyla-
tion is not acting on epitope expression by steric hindrance but rather that
selective carbohydrates behave as molecular effectors of protein conforma-
tion to deliver a variable glycoform pattern. Interestingly, it has been
established that new assays can be developed to optimize the measurement
of a given glycoprotein in blood whether it is a marker or a drug.
Glycoform specific assays may be of invaluable use in providing accurate
tests and assisting clinical trials. Since preparations from human sources are
no longer recognized as available sources of protein antigens, it is crucial to
be able to produce recombinant glycoforms which resemble as closely as
possible to the protein to detect. Glycoengineering is expected to fill the gap
currently existing between the recombinant product and its circulating
(pathological) forms.
1.6 Impact of glycans on production, safety and regulatory aspects

Living cells react quite sensitively to even the slightest changes in their
environment. The combinatorial nature of N-glycosylation provides a
molecular and structural basis to these changes. Many factors, from the
nutrient solution to the design of the equipment affect glycan composition.
For instance, Van Berkel et al. have observed changes in N-glycosylation
from batch production in shaker to fed-batch production in bioreactor [63].
As a consequence, it is quite difficult to keep a given glycoform profile
constant from batch-to-batch. This renders the production highly challen-
ging as it tolerates few to none changes in the manufacturing process during
the lifetime of a drug. Regulatory Agencies such as US Food and Drug
Administration and European Medicines Agency are concerned about
the heterogeneity of preparations and the consistency of bioproduction.
Carbohydr. Chem., 2012, 38, 94–123 | 103

Since each glycoform brings its own pharmacokinetic, pharmacodynamic
and efficacy profile, the activity of resulting mixture is hardly predictable.
Therefore, glycoprotein heterogeneity soon became a major concern as it
poses problems in the quantitative evaluation of the drug principle. The best
approach so far thus appears to minimize glycosylation heterogeneity and
keep it constant throughout development and manufacturing to ensure safety
control and maintenance of both potency and pharmacokinetic properties.
The authorities are currently raising new standards that biomanufacturers
must follow to address these issues and analyze the glycoprofile of their
products [91–93]. Manufacturers are requested to characterize the carbo-
hydrate structure and degree of heterogeneity of their products in quanti-
tative terms in relation to the manufacturing process. It is expected that they
demonstrate comparability between preclinical and clinical batches and
perform comparability assessment to support changes in manufacturing
process during the development of their product and post-approval.
Glycosylation is also an issue in the evolution of biosimilars [15]. Even
with full information about the process, it is not easy to replicate the
manufacturing process to obtain a product similar if not identical to the
original innovator drug. Because of unavoidable differences in bioproduc-
tion, regulators request a high level of scrutiny and often additional clinical
data proving that the new drugs are at least as safe and effective as the
originals. In extreme cases, it can lead to a whole new marketing author-
ization procedure. Genzyme recently had to file another application to show
that Myozyme, a lysosomal glycogen-specific enzyme indicated for the
treatment of patients with Pompe disease, manufactured at a location
different from the original is as safe and effective than the first batch.
Overall, the past decades have unravelled the remarkable complexity of
the glycan repertoire. Extensive characterization of expression systems
revealed that their glycosylation machinery is species specific and differ
from human cells. When produced in these systems, human protein-based
therapeutics often showed a quite variable glycosylation profile that
substantially impact their biological performances. In most cells, the same
glycosylation enzymes are working on a wide panel of endogenous accep-
tors as well as on an overexpressed protein with distinct efficiency. We are
still lacking a full understanding on the mechanisms which regulate in vivo
glycosylation so tightly. To date, variable heterogeneity is regarded as a
mean for each protein to accomplish several functions with variable
intensity. By altering both the folding and/or fate of a given protein, glycans
allows a versatile setting of its specific interactions along with its biological
activity and may thereby also govern variable response at the target
tissue(s). For example, non fucosylated IgG Fc present in human serum
have increased ADCC activity while the presence of sialic acid provides
anti-inflammatory properties to antibodies. In antibodies and also in other
proteins, we know now that both inner and terminal glycosylation appear as
key molecular modulators of glycoprotein structure and activity. It is
therefore important to integrate this recent knowledge in designing glyco-
protein production. Engineering cells to optimize therapeutic proteins and/
or bioprocesses should address the manipulation of the glycoform profile to
reach better detection, efficacy and/or safety of the desired drug.
104 | Carbohydr. Chem., 2012, 38, 94–123

2 Glycosylation optimization
Cell lines and bioproduction today are no longer developed solely on the
basis of cell growth and production yield but have now integrated the need
of controlling N-glycosylation. Culture processes and culture media are
evaluated to identify the conditions that promote optimal glycoform profile
and reproducibility. Mammalian cell lines may be genetically modified for
producing a more homogeneous and non-immunogenic glycosylation.
Alternatively, postproduction glycosylation by in vitro enzymatic reactions
may be used to release undesired terminal sugars and/or complete the oli-
gosaccharide chain enzymatically. Development of new therapeutic glyco-
proteins will therefore include customization of glycosylation, a more
efficient glycan synthesis during bioproduction, and extensive character-
ization of the glycosylation pattern of the product.
2.1 Tuning cell culture conditions to control glycosylation

Since the approval of the first recombinant protein, biomanufacturing
underwent tremendous progress over the past decades [27]. The improve-
ments include optimization of cell lines, development of high-performing
media free of animal-derived components as well as advances in bioreactor
engineering and purification processes. In addition, fundamental studies on
metabolic pathways have shown that culture conditions such as pH and
temperature greatly impact the glycosylation profile. Glycoengineering is
therefore being accomplished first by selecting the clone that produces the
optimal glycoprofile among many overproducer clones for each given
protein [63]. Once the clone is selected, tuning of the culture conditions has
to be performed to favor production of the appropriate glycoforms. Mul-
tiple parameters of cell culture processes and media composition have been
shown to influence cell glycosylation machinery and therefore impact gly-
coform profile [94]. They include medium components for instance metal
ions, butyrate, glucose and Gal concentration, cell culture environment as
well as growth rate and cell productivity [95]. Screening and tuning of these
conditions need to be performed on a case-by-case basis since glycosylation
is protein-specific and because a variable glycan pattern cannot be easily
predicted based on the use of the expression system [63].
Culture conditions affect altogether occupancy of glycosylation sites,
branching and completion of oligosaccharide chains [95–97], presumably by
altering the expression of several glycosylation enzymes which consist in
various nucleotide sugar synthases and transporters, glycosidases and
glycosyltransferases. Expression profiling of 24 N-glycosylation genes in
CHO cells has shown that 21 were significantly up- and down- regulated
during culture phases inducing change in IFN-g N-glycosylation [98].
Addition of nutriments to manipulate nucleotide sugar metabolism has also
been used to improve complete oligosaccharide chain synthesis. Feeding
IFNg-producing CHO fed-batch cultures with glutamine and glucose
increased the content in hybrid and high mannose glycans. Adding galac-
tose was found to facilitate a more galactosylated N-glycan profile [99] and
subsequent addition of sialic acid in recombinant antibodies. Supple-
mentation of CHO culture with N-acetylmannosamine (ManNAc), a
Carbohydr. Chem., 2012, 38, 94–123 | 105

precursor of sialic acid, resulted in increased sialylation of IFN-g [100].
Other components may alter the glycosylation profile by limiting the pre-
sence of different sugars. Decrease in NeuGc content of a therapeutic
antibody expressed by CHO cells can be achieved by simply adding
N-acetylneuraminic acid (NeuAc) to the culture media [71]. Yet, the same
experiment using murine myeloma cell lines did not produce similar data,
probably because the uptake of NeuAc is poorly efficient in this cell type.
Inhibitors have been also used to reduce activities of glycosyltransferases
and favor the synthesis of specific glycans. Zhou et al. used the potent
inhibitor of a-mannosidase, kifunensine, to produce non-fucosylated, oli-
gomannosidic structures [101]. Scientists at Seattle Genetics modified
monoclonal antibody carbohydrates by adding to CHO cultures small
fucose analogs. The fucose analogs inhibit GDP-mannose dehydratase
(GMD), the first enzyme acting in the de novo synthesis of GDP fucose, and
therefore extensively deplete the pool of intracellular GDP-fucose. The
resulting antibodies showed a significant reduction in their carbohydrate
fucosylation [102].
Exoglycosidase activities are other important factors that affect the
glycoform profile of a recombinant glycoprotein post-secretion. During
bioproduction, cell death and lysis may release cytosolic sialidases and
galactosidases which degrade the oligosaccharide chains of newly synthe-
sized proteins [103]. Gu et al have shown a remarkable loss of sialic acids in
IFN-g along with concomitant decrease in CHO cell viability throughout
long-term batch cultivation [104]. Neuraminidase degradation can be
minimized by adding an inhibitor, a substrate analogue i.e 2,3-dehydro-2-
deoxy-NeuAc, to the culture medium of cells expressing a recombinant
antibody [105].
Modification of processes, optimization of media and addition of
supplements can thus exert some control over the protein glycosylation
pathway, favor some glycoforms over others and reduce heterogeneity to
some extent. It is worth noticing that the control over glycosylation is
limited to the pattern constructed by the cell type and there is still a need
for further improvement. An increased knowledge on the regulation of N-
glycosylation enzymes and cell responses during fermentation will certainly
benefit the control of the overall bioproduction process.
2.2 Postproduction modifications

Another approach for remodeling glycosylation has been performed by
in vitro enzymatic modification. This is a postproduction process using
soluble enzymes to reshape glycan structures in vitro and obtain a more
homogeneous preparation of recombinant glycoprotein. Cloning techniques
have made available more than 30 glycosyltransferases from various sources
[106] and deletion of their membrane anchor allowed isolation of soluble
enzymes and further use in enzymatic reactor [107]. It was observed that
in vitro completion of glycans definitely improved performances of
therapeutic glycoproteins, especially their stability and pharmacokinetics.
The process is not cost-effective as it requires quantities of both enzymes
and nucleotide sugar donors and necessitates additional purification steps.
Yet, commercial therapeutic proteins have been generated through this
106 | Carbohydr. Chem., 2012, 38, 94–123

method. Cerezyme, a lysosomal glycoprotein enzyme analog of the human
enzyme b-glucocerebrosidase prescribed for the treatment of Gaucher
disease required postproduction chemical modification for full function.
Cerezyme is produced by recombinant DNA technology using mammalian
cell culture in CHO cells. The oligosaccharide chains have been modified
postproduction to terminate in phosphorylated mannose residues in order
to facilitate mannose receptor-mediated uptake by macrophages [108].
N-glycan remodeling in vitro has been generally be successfully achieved at
the research level [54, 61, 109, 110], especially for terminal glycosylation in
order to increase duration in blood. Such synthesis requires sequential
addition of these residues by several enzymes and the process further involves
rather complex purification procedures to isolate the final product from
unreacted material. Raju et al. have remodeled the oligosaccharide chain of a
Fc domain fused to the extracellular portion of the TNF receptor (TNFR-
Fc), produced in CHO cells, by using a b1,4-galactosyltransferase (GalT) and
an a2,3sialyltransferase (ST3). Terminal sialylation was increased by
approximately 23% preferentially in the TNFR domain rather than in the Fc
domain [109]. More recently, Anthony et al. could sialylate immunoglobulin-
derived Fc in vitro based on the use of a2,3 – (ST3) or a2,6 sialyltransferase
(ST6) following GalT treatment. Interestingly, ST6 efficiency was found to be
more effective than ST3, suggesting that glycan accessibility may affect the
efficiency of these sialyltransferases (STs) [54]. Other enzymes such as
recombinant N-acetylglucosaminyltransferase III (GnTIII) which increases
the content in bisecting GlcNAc have been cloned and tested successfully for
in vitro enzymatic modification of Fc domain and found to improve effector
function [61].
Altogether these studies demonstrated that N-glycans can be remodeled
by using engineered glycosyltransferases. These enzymes possess a good
potential for achieving glycan customization but the methodology is limited
for an industrial biomanufacturing process especially at a GMP scale. The
extra purification steps as well as the cost of goods required by this
approach are definitely unappealing. Manufacturing transferases and acti-
vated sugars at GMP grade is costly and it is probably not a sustainable
solution for production at a commercial scale. In addition, sugar addition
may also be limited by a lack of accessibility of the glycan in fully folded
proteins and can thus never reach completion as during biosynthesis and
thus the needed improvement in biopotency.
2.3 Glycoengineering of rodent cells

A good alternative to enzymatic remodelling of glycoproteins resides in the
development of knockout and knockin cells with glycosyltransferases.
The increasing demand for therapeutic proteins led to the development of
genetic approaches to obtain high expressing cell lines [27] and robust
bioproduction. Cell engineering is routinely employed to modify cellular
processes of host cells and increase protein secretion rate [111]. Identifica-
tion and cloning of glycosyltransferase genes have provided new tools for
further manipulating expression systems [112]. There is currently intensive
work in this field to control sugar addition and improve glycoform profile
and generate protein with better performances.
Carbohydr. Chem., 2012, 38, 94–123 | 107

2.3.1 Engineering core fucosylation. ADCC enhancement is of a great
interest for cancer treatments with antibodies and may have major appli-
cation in antibody drugs against infectious diseases as well [113]. The level
of ADCC depends on the phenotype of the polymorphic FcgIIIa-receptor,
limiting the efficiency of the original antibody to about 30% of the patient
population. ADCC-enhanced antibodies are expected to treat new patient
subpopulations presenting a FcyIIIa receptor mutant. They will also allow a
reduction of effective dosage for all patients and thereby reduce side effects.
As described above, altering Fc-linked glycans impacts significantly IgG
effector functions through core fucose, bisecting GlcNAc and/or branching.
The development of glycoengineered cell lines to produce glycoform of
antibodies with increased ADCC function is currently under intensive
exploitation and the first glycoengineered antibody has gained approval in
march 2012.
The first approach to increase ADCC in antibody-based therapeutics is to
equip cells with the enzyme that adds bisecting GlcNAc to N-linked glycans.
The enzyme GnTIII that catalyzes the transfer of a b1,4-GlcNAc residue to
the core b-mannose residue residue is lacking in CHO, NS0 and Sp2/0 cell
lines. Davies et al. from Idec Pharmaceuticals/Biogen co-expressed rat
GnTIII in a CHO/DG44 production cell line along with the anti-CD20
Rituximab (trade name Rituxans and MabTheras) and showed an increase
in up to 20-fold of ADCC activity compared to that of Rituxans produced in
wild type CHO cells. An expression level of the enzyme was sufficient to
ensure glycan branching in the antibody but did not affect cell growth in
suspension in serum- and protein-free cell culture medium nor antibody
production [59]. Scientists at GlycArt overexpressed the rat GnTIII enzyme,
engineered with different localization domains of Golgi-resident enzymes in
HEK293-EBNA cells [114] (Fig. 3a). The antibodies produced were enriched
in bisected non-fucosylated oligosaccharides. GnTIII activity appears to
inhibit core fucosylation suggesting that lack of fucose rather than the pre-
sence of bisected GlcNAc enhanced ADCC. The authors suggested that the
synthesis of bisected oligosaccharides is due to an efficient competition of
chimeric GnTIII with endogeneous a1,6-FUT (FUT8) and mannosidase II
rather than a high expression level of the branching enzyme itself. The
delocalization of GnTIII within early Golgi compartments lead to a more
efficient competition with the processing enzymes. The technology named
GlycoMAbs was adapted in a CHO cell line to produce the GA101 antibody
currently in clinical trials [115].
In contrast to most N-linked glycoproteins, over 90% of antibodies
produced in CHO cells are fucosylated. Since non fucosylated IgG1 anti-
bodies have been shown to display enhanced ADCC for anti-cancer anti-
bodies [116], it is highly desirable to produce antibodies without fucose.
Mammalian cells including CHO cells attach a fucose residue to the
innermost GlcNAc residue of the N-linked oligosaccharide via a FUT8
fucosyltransferase, which transfers fucose from GDP-fucose in a a1,6
linkage. Yamane-Ohnuki et al. have generated stable CHO/DG44 cells
producing non fucosylated antibodies by disrupting both FUT8 alleles. The
non fucosylated form showed in vitro a B100 fold increase in the efficacy
of FcgRIII-mediated ADCC compared to the fucosylated product [117].
108 | Carbohydr. Chem., 2012, 38, 94–123

Fig. 3 (a) Schema of wild type rat GnTIII and GnTIII chimeric enzymes constructed by
substituting the natural N-terminal domain by the localization domains of various Golgi-
resident glycosylation enzymes (114] (b) Schema of minigenes delivering non natural human
ST6 enzymes with optimized catalytic activity. The library is based on the combinatorial
assembly of cytoplasmic, transmembrane and stem regions of various glycosyltransferases as
reported in the footnote.
As shown by scientists at Genentech, enhanced ADDC activity of the non

fucosylated trastuzumab (Trade name Herceptins) observed in vitro
translates into in vivo efficacy for solid tumor targeting [118].
Other genes to produce non-fucosylated antibodies from mammalian
cells could be targeted such as GDP-mannose 4,6-dehydratase (GMD),
GDP-6-deoxy-D-lyxo-4-hexulose reductase (RMD) as well as the GDP-
fucose transporter (GFT) [119–122]. Scientists at Kyowa have shown that
GMD knocked-out CHO/DG44 cells is a viable strategy for the production
of very efficient non fucosylated therapeutic antibodies. The GlymaxX
technology proposed by ProBioGen AG uses the expression of the pro-
karyotic enzyme RMD in CHO cells to interrupt the fucose biosynthesis
pathway by capturing a reactive sugar intermediate [121]. A team at Chugai
Carbohydr. Chem., 2012, 38, 94–123 | 109

Pharmaceutical disrupts the GFT gene by sequential homologous recom-
bination in CHO/DXB11 cells. The engineered cells produced more than
93% of non-fucosylated anti-CD317 mAb [122].
Conversely, sialylated Fc glycans have been shown to negatively influence
ADCC antibody function by decreasing affinity of IgG Fc with the FcgIIIa
receptor. Despite all the advantages brought by the presence of sialic
acids as terminal glycan residues, it may be, desirable to reduce terminal
addition of sialic acid in some particular cases which require blockade of
inflammatory cascades. Naso et al. have generated a CHO cell line
expressing both sialidase A and recombinant antibody. Expression in the
condition medium of sialidase A was sufficient to remove sialic acids from
secreted antibodies [123].
Glycoengineered therapeutic antibodies with improved efficacy are
considered as very promising. Another advantage to the increased potency
of these products is the absence of adverse effects since their glycan pattern
is very similar to natural human serum IgGs [22]. Engineered CHO cell
lines to produce such glycoengineered proteins are now commercially
available.
2.3.2 Recent progress in engineering sialylation. Although terminal

glycosylation is definitely a key step for the lifespan of glycosylated proteins,
little progress has been obtained in engineering sialylation in animal cell
lines over the past years. Sialic acid is naturally missing in yeast, algae,
plant, insect and avian cells. When expressed in yeasts glycoengineered to
achieve complex glycosylation [124], significant transfer of sialic acid has
been obtained in selected strains but fully sialylated glycans has not been
demonstrated. It is also worth noting that sialic acid is poorly expressed in
mammalian cells which terminate their glycans mostly with a-linked Gal.
Indeed, CHO cells have originally been selected because their a1,3 galac-
tosyltransferase is mutated in its catalytic domain and cannot add a-linked
Gal to glycoproteins . These cells however do express a ST3 activity and are
able to transfer sialic acid as NeuAc and NeuGc to complex glycans in the
a-2,3-linkage. However, the level of sialic acid in glycoproteins produced in
these cells may be artefactually low because of endogeneous sialidases
released in the cell culture medium during production. Several groups have
tackled this issue by developing CHO cell lines in which the sialidase gene
has been silenced [125–127]. A significant reduction in asialoforms has been
observed. It is worth noting also that the human retinoblastoma cell line
PER C6 has been reported to share a similar glycan pattern as CHO and
NSO cells for antibody production [128]. Overall, it therefore appears that
to date, no expression system routinely used for drug production is able to
achieve a2,6-sialylation.
This type of glycosylation is often regarded as human-like glycosylation
because, contrary to other mammals, circulatory human proteins are pro-
minently terminated in a2,6-linked sialic acid. It may indeed be considered
as a signal which prevents the protein of interest from uptake from elim-
ination tissues and from recognition by the immune system because sialic
acid is crucial for absorption, serum half-life and metabolic clearance.
Considering that more than 70% of the marketed proteins are glycosylated,
110 | Carbohydr. Chem., 2012, 38, 94–123

the availability of cells expressing an efficient a2,6-linked sialyltransferase
has long been known as a major goal to achieve.
Engineering efficient sialylation activity is not a straightforward
process. Several strategies based on genetic engineering of rodent cell
lines have been followed to increase sialic acid content of therapeutic pro-
teins. They include addition of glycan acceptor sites in the protein drug,
inhibition of sialidase activities and overexpression of STs with improved
efficacy [129].
To increase the number of sialylation acceptor sites, Fukuta et al.
increased the content in multi-antennary sugar chains on IFN-g by stable
overexpression of N-acetylglucosaminyltransferase IV and V (GnTIV and
GnTV) [130]. However, sialylation did not increase proportionally with the
number of acceptor sites probably because of insufficient activity of either
STs and/or CMP-sialic acid transporters. In further studies, the authors
further extended the method by overexpressing mouse ST3 and/or rat ST6
in CHO cells stably transfected with GnTV. Sialylation activities was
increased by approximately 80% to reach 61.2% in a2,3-and a2,6-linked
sialic acid content, indicating a limited availability of the sialylation
enzymes in CHO cells [131].
Increasing sialylation acceptor sites can also be achieved by increasing
galactosylation efficiency. Weikert et al. have decreased TNK-tPA and
TNF-Fc glycoform heterogeneity by increasing sialylated glycoforms using
CHO cell lines overexpressing human b1,4-GalT and/or ST3. Over 90% of
the resulting glycans were found to be capped with sialic acid [132].
Efficient sialylation does not seem to rely on the presence of a STs only as
CHO transfection with ST3 alone hardly changed the content in sialic acid
in rhEPO compared to production in CHO cotransfected with b1,4-GalT
and ST3 [29]. Rather, increased sialic acid transfer appears also related to
the availability of the CMP-sialic acid, the donor substrate for STs within
the Golgi lumen. Coexpression of CMP-sialic acid transporter (CMP-SAT),
CMP-sialic acid synthase together with human ST3, greatly improved sia-
lylation of rhEPO produced in CHO cells [133]. Another approach to
increase availability of CMP-sialic acid is to increase its biosynthesis in the
cytoplasm. The enzyme UDP-N-acetylglucosamine 2-epimerase/N-acet-
ylmannosamine kinase (GNE/MNK) is a key enzyme in the biosynthesis
process of CMP-sialic acid but is negatively regulated by CMP-sialic acid
[134]. Overexpression of a mutant form of GNE/MNK found in sialuria
disorder has been shown to increase the intracellular pool of sialic acids and
consequently increase sialylation of rhEPO because the feedback mechan-
ism was abolished [135]. Highly sialylated therapeutic glycoproteins were
indeed obtained through overexpression of sialuria-mutated GNE/MNK,
ST3 and CMP-SAT [136].
Since the drug approved cell lines are all missing a ST6 enzyme [137],
several groups have tried to equip rodent cells with ST6 enzymes. The first
introduction of rat ST6GalI in CHO cells was reported by Lee and col-
leagues in 1989 [137]. Rat or human ST6-expressing CHO cells were suc-
cessively generated and tested for various therapeutic glycoproteins
production [138–141]. In all cases, the content in sialic acid did not rise
tremendously and the ratio of a2,6-to a2,3-linked sialic acid never exceeded
Carbohydr. Chem., 2012, 38, 94–123 | 111

50%. It was concluded at the time that the recombinant ST6 enzymes
poorly competes with the endogenous ST3 for the same donor and accep-
tors. Such ratio also indicated a lack of completion in terminal glycosylation
and therefore precluded further industrial development.
Earlier work from our group showed that sialic acid transfer in the 6
position may also be limited because of an exquisite specificity of the ST6
enzyme towards its protein acceptors. The wild type ST6GalI was found to
exhibit high preference for triantennary glycans and reduced activity on
biantennary and tetraantennary glycans. Deleting the enzyme in its N-
terminal half was shown to abolish such specificity and improve sialic acid
transfer by 30-fold [142]. When transfected into CHO cells, such truncated
Fig. 4 Overview of the two main glycoengineering strategies in mammalian expression systems
for modifying the biantennary oligosaccharide of the Fc fragment of human IgG. The size of
each carbohydrate residue is representative of its abundance in the final product. Symbols:
Square represents N-acetylglucosamine, triangle core fucose, circle mannose, circle galactose,
diamond N-glycolylneuraminic (NeuGc) or N-acetylneuraminic acid (NeuAc).
112 | Carbohydr. Chem., 2012, 38, 94–123

variants advantageously located in Golgi compartments involved in the
secretory pathway, in a place where they can meet protein acceptors traf-
ficking to the cell surface [143]. It therefore appeared that ST6 enzymes
require both optimization of their catalytic properties and an appropriate
targeting to the secretory compartments of CHO cells to exhibit optimal
enzymatic activity. Subsequently, ST6 minigenes have been constructed
(Fig. 3b) to combine the optimized catalytic domain with a series of
different cytoplasmic, transmembrane and stem regions and obtain non
natural enzymes highly active in vivo (see footnote). The company Sia-
Med’Xpress is exploiting the transferase minigenes to equip a wide array of
cell lines and significantly enhance sialylation of any glycoprotein of interest
with high efficiency. Such optimized glycosyltransferases can be adapted to
any drug family to achieve optimal sialic acid transfer under the best cell
culture conditions. This presently represents a major advance in manip-
ulating terminal glycosylation in drug approved cell lines.
Finally, engineering terminal sialylation may advantageously be com-
bined with existing methods to achieve for the first time a full panel of
glycoengineered cells to provide a consistent and complete pattern of
human-like glycosylation.
3 Conclusions
A glycosylation-based strategy to optimize protein drugs is now available
and schematized in Fig. 4. For a long time, sialylation engineering has been
poorly efficient but the design of highly active transferases has recently
opened the wide prospective of tailored glycosylation in drug approved cell
lines. Many expression systems including human cells themselves have been
extensively characterized to this purpose and each of them presents its own
advantages and delivers a human protein with a content in glycoforms
significantly different from that of the natural protein. At present, the
overall methodology is thus well suited to respond to the current therapeutic
needs of many new candidates. Methods have now been developed to
produce antibodies with customized glycosylation and improved cytotoxi-
city and these molecules pioneer the field of glycoengineering.
At present, glycoengineering is an advanced technology which may
improve production yield as glycosylation is known to (i) increase secretion
[144], (ii) reduce misfolding/aggregation [3, 30] and (iii) enhance bioactivity
as sialic acid is a key for prolonged duration in the circulation [33].
Meanwhile, it will help reducing overall costs as bioproduction will be more
efficient in delivery products in higher yield, better stability, reduced
immunogenicity and improved activity. It will also substantially reduce
healthcare costs through enhanced bioavailability and low dosage of the
desired drug.
Because glycosylation is species, protein and cell-specific, achieving drug
optimization through glycoengineering is an exciting challenge. It is clear
The ‘‘Preparation and Uses of Gene Sequences encoding Chimerical Glycosyltransferases

with Optimized Glycosylation Activity’’ has been patented as EP2019864 [C. Ronin and
G. Guiraudie-Capraz, inv].
Carbohydr. Chem., 2012, 38, 94–123 | 113

that a glycosylation-based strategy needs to start early in the development
stage to optimize safety and efficacy profile of each protein candidate. For
that purpose, expression systems should be selected as to provide increased
secretion along with proteins with reduced immunogenicity and improved
glycoform patterns. Manipulating cell culture conditions will further opti-
mize the glycoform profile to meet the regulatory requirements and give the
best opportunity to select lead candidates. Purification will finally select the
highly sialylated forms which will ensure enhanced blood duration and
improve efficacy during further drug development. Meanwhile, desired
glycoform pattern and bioactivity testing will be screened along with the
strategy to finally reach the best profile for drug efficacy. It may thus be
anticipated that at each stage of development, glycoengineering will take
place differently to achieve final optimal improvement in drug perfor-
mances. Indeed, bioproduction of glycoengineered proteins needs to be
evaluated on a case-by-case basis. Such multistep strategy will have to be
adapted to each protein from the very beginning and will hopefully over-
come most of the difficulties encountered so far, thereby speeding up the
scale up production and limiting costs at each step.
Of special interest, accurate estimation of glycosylated biomarker or drug
amount is of paramount importance throughout glycoform optimization.
Valuable measurement, especially during pharmacokinetic studies, is the
key in assessing efficacy improvement. However, a number of current
immunoassays heavily discriminate glycoforms because they differ in
immunological properties. Quantitative measurement of a human protein
produced by recombinant technology may be greatly hampered by epitope
non-similarity between preparations of natural and recombinant origin.
Consequently, measuring a combination of glycoforms and searching for
a gain in bioactivity may readily lack accuracy, especially in blood [76].
While structural analysis is of definite help in characterizing the wide
panel of glycans, it does not always estimate the ratio of the various
glycoforms accurately. There is thus a clear need to achieve quantitative
assessment of glycoform mixture at all steps of drug development, including
clinical trials. Development of new assays based on glycoengineered
proteins and glycoform-specific antibodies is the most promising advance at
the moment.
All the knowledge acquired so far in the field of therapeutic proteins will
be of great help to identify the key steps that can be specifically targeted for
glycoengineering. To date, terminal sialylation, in most proteins, and core
fucosylation, typically in anti-cancer antibodies, are the two major effectors
that have been identified among marketed drugs. The remarkable knowl-
edge issued from high throughput glycome analysis will help identifying the
proper changes in glycosylation to reach therapeutic targets and/or restore
normal function in diseased or age-related states. Focusing glycoengineer-
ing on these key issues over a new array of different therapeutic proteins will
undoubtedly promote the design of new drugs. Meanwhile, it will also
demonstrate that the glycan repertoire can also be advantageously
manipulated for the benefit of a wide array of protein targets. Over time,
molecular and cellular engineering may be able to design and produce by
the key molecules of many different biomedical applications.
114 | Carbohydr. Chem., 2012, 38, 94–123

References
1 R. G. Spiro, Glycoproteins, AdvProteinChem, 1973, 27, 349–467.
2 H. Lis and N. Sharon, Protein glycosylation. Structural and functional
aspects, Eur J Biochem, 1993, 218, 1–27.
3 R. J. Solá and K. Griebenow, Glycosylation of therapeutic proteins: an
effective strategy to optimize efficacy, Bio Drugs, 2010, 24, 9–21.
4 P. M. Rudd, T. Elliott, P. Cresswell, I. A. Wilson and R. A. Dwek, Glyco-
sylation and the immune system, Science, 2001, 291, 2370–6.
5 K. Ohtsubo and J. D. Marth, Glycosylation in cellular mechanisms of health
and disease, Cell, 2006, 126, 855–67.
6 J. Wang, C. I. Balog, K. Stavenhagen, C. A. Koeleman, H. U. Scherer, M. H.
Selman, A. M. Deelder, T. W. Huizinga, R. E. Toes and M. Wuhrer,
Fc-glycosylation of IgG1 is modulated by B-cell stimuli, Mol Cell Proteomics,
2011, 10.
7 J. Zhao, D. M. Simeone, D. Heidt, M. A. Anderson and D. M. Lubman,
Comparative serum glycoproteomics using lectin selected sialic acid glyco-
proteins with mass spectrometric analysis: application to pancreatic cancer
serum, J Proteome Res, 2006, 5, 1792–802.
8 A. Ercan, J. Cui, D. E. Chatterton, K. D. Deane, M. M. Hazen, W. Brintnell, C. I.
O’Donnell, L. A. Derber, M. E. Weinblatt, N. A. Shadick, D. A. Bell, E. Cairns,
D. H. Solomon, V. M. Holers, P. M. Rudd and D. M. Lee, Aberrant IgG
galactosylation precedes disease onset, correlates with disease activity, and is
prevalent in autoantibodies in rheumatoid arthritis, Arthritis Rheum, 2010, 62,
2239–48.
9 Y. Sato and T. Endo, Alteration of brain glycoproteins during aging, Geriatr
Gerontol Int, 2010, 10, 32–40.
10 E. Reynders, F. Foulquier, W. Annaert and G. Matthijs, How Golgi glyco-
sylation meets and needs trafficking: the case of the COG complex, Glyco-
biology, 2011, 21, 853–63.
11 E. Lisowska, The role of glycosylation in protein antigenic properties, Cell
Mol Life Sci, 2002, 59, 445–55.
12 A. M. Sinclair and S. Elliott, Glycoengineering: the effect of glycosylation on
the properties of therapeutic proteins, J Pharm Sci, 2005, 94, 1626–35.
13 R. Jefferis, Glycosylation of recombinant antibody therapeutics, Biotechnol
Prog, 2005, 21, 11–6.
14 G. Walsh and R. Jefferis, Post-translational modifications in the context of
therapeutic protein, Nature Biotechn, 2006, 24, 1241–52.
15 N. Kawasaki, S. Itoh, N. Hashii, D. Takakura, Y. Qin, X. Huang and T.
Yamaguchi, The significance of glycosylation analysis in development of
biopharmaceuticals, Biol Pharm Bull, 2009, 32, 796–800.
16 H. Li and M. d’Anjou, Pharmacological significance of glycosylation in
therapeutic proteins, Curr Opin Biotechnol, 2009, 20, 678–84.
17 U. H. Stenman, Standardization of assays for human chorionic gonadotropin,
Clin Chem, 2004, 50, 798–800.
18 J. R. Moskal, R. A. Kroes and G. Dawson, The glycobiology of brain tumors:
disease relevance and therapeutic potential, Expert Rev Neurother, 2009, 9,
1529–45.
19 P. M. Drake, W. Cho, B. Li, A. Prakobphol, E. Johansen, N. L. Anderson, F. E.
Regnier, B. W. Gibson and S. J. Fisher, Sweetening the pot: adding glycosy-
lation to the biomarker discovery equation, Clin Chem, 2010, 56, 223–36.
20 F. M. Wurm, Production of recombinant protein therapeutics in cultivated
mammalian cells. Nat Biotechnol, 2004, 22, 1393–8.
Carbohydr. Chem., 2012, 38, 94–123 | 115

21 Y. Durocher and M. Butler, Expression systems for therapeutic glycoprotein
production, Curr Opin Biotechnol, 2009, 20, 700–7.
22 R. Jefferis, Glycosylation as a strategy to improve antibody-based ther-
apeutics, Nat Rev Drug Discov, 2009, 8, 226–34.
23 J. R. Rich and S. G. Withers, Emerging methods for the production of
homogeneous human glycoproteins, Nat Chem Biol, 2009, 5, 206–15.
24 M. Bencúrová, W. Hemmer, M. Focke-Tejkl, I. B. Wilson and F. Altmann,
Specificity of IgG and IgE antibodies against plant and insect glycoprotein
glycans determined with artificial glycoforms of human transferrin, Glyco-
biology, 2004, 14, 457–66.
25 S. Hino, F. Umeda, S. Inumaru, N. Aoki, C. Sato, T. Okajima, D. Nadano
and T. Matsuda, IgG2 dominancy and carbohydrate recognition specificity of
C3H/He mouse antibodies directed to cross-reactive carbohydrate determi-
nants (CCDs) bearing beta-(1,2)-xylose and alpha-(1,3)-fucose, Immunol Lett,
2010, 133, 28–34.
26 S. R. Hamilton and T. U. Gerngross, Glycosylation engineering in yeast: the
advent of fully humanized yeast, Curr Opin Biotechnol, 2007, 18, 387–92.
27 Y. Lim, N. Wong, Y. Y. Lee, S. Ku, D. C. Wong and M. G. Yap, Engineering
mammalian cells in bioprocessing - current achievements and future perspec-
tives, Biotechnol Appl Biochem, 2010, 55, 175–89.
28 W. Morelle, S. Donadio, C. Ronin and J. C. Michalski, Characterization of N-
glycans of recombinant human thyrotropin using mass spectrometry, Rapid
Commun Mass Spectrom, 2006, 20, 331–45.
29 Y. T. Jeong, O. Choi, H. R. Lim, Y. D. Son, H. J. Kim and J. H. Kim,
Enhanced sialylation of recombinant erythropoietin in CHO cells by human
glycosyltransferase expression, J Microbiol Biotechnol, 2008, 18, 1945–52.
30 X. Xu, H. Nagarajan, N. E. Lewis, S. Pan, Z. Cai, X. Liu, W. Chen, M. Xie,
W. Wang, S. Hammond, M. R. Andersen, N. Neff, B. Passarelli, W. Koh, H.
C. Fan, J. Wang, Y. Gui, K. H. Lee, M. J. Betenbaugh, S. R. Quake, I. Famili,
B. O. Palsson and J. Wang, The genomic sequence of the Chinese hamster
ovary (CHO)-K1 cell line, Nat Biotechnol, 2011, 29, 735.
31 G. Vegarud and T. B. Christnsen, Glycosylation of Proteins: a new method of
enzyme stabilization, Biotechnol Bioeng, 1975, 17, 1391–7.
32 M. Baudys, T. Uchio, D. Mix, D. Wilson and S. W. Kim, Physical stabiliza-
tion of insulin by glycosylation, J Pharm Sci, 1995, 84, 28–33.
33 V. Kayser, N. Chennamsetty, V. Voynov, K. Forrer, B. Helk and B. L. Trout,
Glycosylation influences on the aggregation propensity of therapeutic mono-
clonal antibodies, Biotechnol J, 2011, 6, 38–44.
34 T. S. Raju and B. Scallon, Fc glycans terminated with N-acetylglucosamine
residues increase antibody resistance to papain, BiotechnolProg, 2007, 23,
964–71.
35 A. J. Jones, D. I. Papac, E. H. Chin, R. Keck, S. A. Baughman, Y. S. Lin, J.
Kneer and J. E. Battersby, Selective clearance of glycoforms of a complex
glycoprotein pharmaceutical caused by terminal N-acetylglucosamine is
similar in humans and cynomolgus monkeys, Glycobiology, 2007, 17, 529–40.
36 A. G. Morell, G. Gregoriadis, I. H. Scheinberg, J. Hickman and G. Ashwell,
The role of sialic acid in determining the survival of glycoproteins in the cir-
culation, J BiolChem, 1971, 246, 1461–7.
37 L. M. Steirer, E. I. Park, R. R. Townsend and J. U. Baenziger, The asia-
loglycoprotein receptor regulates levels of plasma glycoproteins terminating
with sialic acid alpha2,6-galactose, J BiolChem, 2009, 284, 3777–83.
38 D. W. Briggs, J. W. Fisher and W. J. George, Hepatic clearance of intact and
desialylated erythropoietin, Am J Physiol, 1974, 227, 1385–8.
116 | Carbohydr. Chem., 2012, 38, 94–123

39 I. C. Macdougall, Optimizing the use of erythropoietic agents--pharmacokinetic
and pharmacodynamic considerations, Nephrol Dial Transplant, 2002, 17, 66–70.
40 B. Byrne, G. G. Donohoe and R. O’Kennedy, Sialic acids: carbohydrate
moieties that influence the biological and physical properties of biopharma-
ceutical proteins and living cells, Drug Discov Today, 2007, 12, 319–26.
41 J. C. Egrie, E. Dwyer, J. K. Browne, A. Hitz and M. A. Lykos, Darbepoeti-
nalfa has a longer circulating half-life and greater in vivo potency than
recombinant human erythropoietin, ExpHematol, 2003, 31, 290–9.
42 C. Wacker, C. N. Berger, P. Girard and R. Meier, Glycosylation profiles of
therapeutic antibody pharmaceuticals, Eur J Pharm Biopharm, 2011, 2.
43 A. W. Guzzetta, L. J. Basa, W. S. Hancock, B. A. Keyt and W. F. Bennett,
Identification of carbohydrate structures in glycoprotein peptide maps by the
use of LC/MS with selected ion extraction with special reference to tissue
plasminogen activator and a glycosylation variant produced by site directed
mutagenesis, Anal Chem, 1993, 1, 2953–62.
44 D. P. Beebe and D. L. Aronson, Turnover of tPA in rabbits: influence of
carbohydrate moieties, Thromb Res, 1988, 51, 11–22.
45 M. Einarsson, J. Brandt and L. Kaplan, Large-scale purification of human-
tissue-type plasminogen activator using monoclonal antibodies, Biochim Bio-
phys Acta, 1985, 830, 1–10.
46 T. S. Raju, Terminal sugars of Fcglycans influence antibody effector functions
of IgGs, Curr Opin Immunol, 2008, 20, 471–8.
47 A. Wright and S. L. Morrison, Effect of glycosylation on antibody function:
implications for genetic engineering, Trends Biotechnol, 1997, 15, 26–32.
48 J. M. Woof and D. R. Burton, Human antibody-Fc receptor interactions
illuminated by crystal structures, Nat Rev Immunol, 2004, 4, 89–99.
49 R. B. Parekh, R. A. Dwek, B. J. Sutton, D. L. Fernandes, A. Leung, D.
Stanworth, T. W. Rademacher, T. Mizuochi, T. Taniguchi, K. Matsuta, F.
Takeuchi, Y. Nagano, T. Miyamoto and A. Kobata, Association of rheu-
matoid arthritis and primary osteoarthritis with changes in the glycosylation
pattern of total serum IgG, Nature, 1985, 1–7, 452–7.
50 Y. Kaneko, F. Nimmerjahn and J. V. Ravetch, Anti-inflammatory activity of
immunoglobulin G resulting from Fc sialylation, Science, 2006, 313, 670–3.
51 J. Bayry, K. Bansal, M. D. Kazatchkine and S. V. Kaveri, DC-SIGN and
alpha2,6-sialylated IgG Fc interaction is dispensable for the anti-inflammatory
activity of IVIg on human dendritic cells, Proc Natl Acad Sci U S A, 2009, 106.
52 A. E. Bedwell, C. J. Elson, S. D. Carter, P. A. Dieppe, C. W. Hutton and R.
Czudek, Isolation and analysis of complement activating aggregates from
synovial fluid of patients with rheumatoid arthritis using monoclonal anti-C3d
antibodies, Ann Rheum Dis, 1987, 46, 55–64.
53 Y. Kaneko, F. Nimmerjahn and J. V. Ravetch, Anti-inflammatory activity of
immunoglobulin G resulting from Fc sialylation, Science, 2006, 313, 670–3.
54 R. M. Anthony, F. Nimmerjahn, D. J. Ashline, V. N. Reinhold, J. C. Paulson
and J. V. Ravetch, Recapitulation of IVIG anti-inflammatory activity with a
recombinant IgG Fc, Science, 2008, 320, 373–6.
55 R. M. Anthony and J. V. Ravetch, A novel role for the IgG Fc glycan: the
anti-inflammatory activity of sialylatedIgG Fcs, J Clin Immunol, 2010, 30, 9–14.
56 J. F. Séı̈té, D. Cornec, Y. Renaudineau, P. Youinou, R. A. Mageed and S.
Hillion, IVIg modulates BCR signaling through CD22 and promotes apop-
tosis in mature human B lymphocytes, Blood, 2010, 116, 1698–704.
57 D. N. Forthal, J. S. Gach, G. Landucci, J. Jez, R. Strasser, R. Kunert and H.
Steinkellner, Fc-glycosylation influences Fcg receptor binding and cell-mediated
anti-HIV activity of monoclonal antibody 2G12, J Immunol, 2010, 1, 6876–82.
Carbohydr. Chem., 2012, 38, 94–123 | 117

58 P. Umaña, J. Jean-Mairet, R. Moudry, H. Amstutz and J. E. Bailey,
Engineered glycoforms of an anti neuroblastoma IgG1 with optimized anti-
body-dependent cellular cytotoxic activity, Nat Biotechnol, 1999, 17, 176–80.
59 J. Davies, L. Jiang, L. Z. Pan, M. J. LaBarre, D. Anderson and M. Reff,
Expression of GnTIII in a recombinant anti-CD20 CHO production cell line:
Expression of antibodies with altered glycoforms leads to an increase in
ADCC through higher affinity for FC gamma RII, Biotechnol Bioeng, 2001,
20, 288–94.
60 C. Ferrara, S. Grau, C. Jäger, P. Sondermann, P. Brünker, I. Waldhauer, M.
Hennig, A. Ruf, A. C. Rufer, M. Stihle, P. Umaña and J. Benz, Unique car-
bohydrate-carbohydrate interactions are required for high affinity binding
between Fc{gamma}RIII and antibodies lacking core fucose, Proc Natl Acad
Sci U S A, 2011, 2, 12669–74.
61 J. Hodoniczky, Y. Z. Zheng and D. C. James, Control of recombinant
monoclonal antibody effector functions by Fc N-glycan remodeling in vitro,
Biotechnol Prog, 2005, 21, 1644–52.
62 P. N. Boyd, A. C. Lines and A. K. Patel, The effect of the removal of sialic
acid, galactose and total carbohydrate on the functional activity of Campath-
1H, Mol Immunol, 1995, 32, 1311–8.
63 P. H. Van Berkel, J. Gerritsen, G. Perdok, J. Valbjørn, T. Vink, J. G. Van de
Winkel and P. W. Parren, N-linked glycosylation is an important parameter
for optimal selection of cell lines producing biopharmaceutical human IgG,
Biotechnol Prog, 2009, 25, 244–51.
64 Y. Kanda, T. Yamada, K. Mori, A. Okazaki, M. Inoue, K. Kitajima-Miyama,
R. Kuni-Kamochi, R. Nakano, K. Yano, S. Kakita, K. Shitara and M. Satoh,
Comparison of biological activity among nonfucosylated therapeutic IgG1
antibodies with three different N-linked Fc oligosaccharides: the high-man-
nose, hybrid, and complex types, Glycobiology, 2006, 17, 104–18.
65 D. Ternant, A. Aubourg, C. Magdelaine-Beuzelin, D. Degenne, H. Watier, L.
Picon and G. Paintaud, Infliximab pharmacokinetics in inflammatory bowel
disease patients, Ther Drug Monit, 2008, 30, 523–9.
66 J. G. Gribben, S. Devereux, N. S. Thomas, M. Keim, H. M. Jones, A. H.
Goldstone and D. C. Linch, Development of antibodies to unprotected
glycosylation sites on recombinant human GM-CSF, Lancet, 1990, 24, 434–7.
67 S. Porter, Human immune response to recombinant human proteins, J Pharm
Sci, 2001, 90, 1–11.
68 N. Casadevall, Antibodies against rHuEPO: native and recombinant, Nephrol
Dial Transplant, 2002, 17, 42–7.
69 A. Noguchi, C. J. Mukuria, E. Suzuki and M. Naiki, Immunogenicity of
N-glycolylneuraminicacid-containing carbohydrate chains of recombinant
human erythropoietin expressed in Chinese hamster ovary cells, J Biochem,
1995, 117, 59–62.
70 A. Zhu and R. Hurst, Anti-N-glycolylneuraminic acid antibodies identified in
healthy human serum, Xenotransplantation, 2002, 9, 376–81.
71 D. Ghaderi, R. E. Taylor, V. Padler-Karavani, S. Diaz and A. Varki, Impli-
cations of the presence of N-glycolylneuraminic acid in recombinant ther-
apeutic glycoproteins, Nat Biotechnol, 2010, 28, 863–7.
72 C. K. Borrebaeck, A. C. Malmborg and M. Ohlin, Does endogenous glyco-
sylation prevent the use of mouse monoclonal antibodies as cancer ther-
apeutics?, Immunol Today, 1993, 14, 477–9.
73 U. Galili, F. Anaraki, A. Thall, C. Hill-Black and M. Radic, One percent of
human circulating B lymphocytes are capable of producing the natural anti-
Gal antibody, Blood, 1993, 15, 2485–93.
118 | Carbohydr. Chem., 2012, 38, 94–123

74 C. H. Chung, B. Mirakhur, E. Chan, Q. T. Le, J. Berlin, M. Morse, B. A.
Murphy, S. M. Satinover, J. Hosen, D. Mauro, R. J. Slebos, Q. Zhou, D.
Gold, T. Hatley, D. J. Hicklin and T. A. Platts-Mills, Cetuximab-induced
anaphylaxis and IgE specific for galactose-alpha-1,3-galactose, N Engl J Med,
2008, 13, 1109–17.
75 G. Antonelli, Development of neutralizing and binding antibodies to inter-
feron (IFN) in patients undergoing IFN therapy, Antiviral Res, 1994, 24,
235–44.
76 S. Donadio, W. Morelle, A. Pascual, R. Romi-Lebrun, J. C. Michalski and
C. Ronin, Both core and terminal glycosylation alter epitope expression in
thyrotropin and introduce discordances in hormone measurements, Clin Chem
Lab Med, 2005, 43, 519–30.
77 C. A. Spencer, M. Takeuchi and M. Kazarosyan, Current status and perfor-
mance goals for serum thyrotropin (TSH) assays, Clin Chem, 1996, 42, 140–5.
78 W. G. Wood, D. Waller and U. Hantke, An evaluation of six solid-phase
thyrotropin (TSH) kits, J Clin Chem Clin Biochem, 1985, 23, 461–71.
79 P. Carayon, E. Martino, L. Bartalena, L. Grasso, C. Mammoli, S. Costagliola
and A. Pinchera, Clinical usefulness and limitations of serum thyrotropin
measurement by ‘ultrasensitive’ methods. Comparisons of five kits, Horm Res,
1987, 26, 105–17.
80 B. Y. Chan and R. Swaminathan, Analytical and clinical performance of six
sensitive thyrotrophin kits, Pathology, 1990, 22, 11–5.
81 P. Laurberg, Persistent problems with the specificity of immunometric TSH
assays, Thyroid, 1993, 3, 279–83.
82 C. A. Spencer and C. C. Wang, Thyroglobulin measurement. Techniques, clinical
benefits, and pitfalls, Endocrinol Metab Clin North Am, 1995, 24, 841–63.
83 W. G. Wood, U. Bruns, O. Eber, W. Langsteger, J. Y. Bounaud and M. P.
Bounaud, Evaluation of the Abbott IMx ultrasensitive II hTSH immuno-
metric assay in three European centers: a comparison with established com-
mercial immunometric assays for thyrotropin, Eur J Clin Chem Clin Biochem,
1996, 34, 151–8.
84 A. K. Rasmussen, L. Hilsted, H. Perrild, E. Christiansen, K. Siersbaek-Nielsen
and U. Feldt-Rasmussen, Discrepancies between thyrotropin (TSH) mea-
surement by four sensitive immunometric assays, Clin Chim Acta, 1997, 259,
117–28.
85 A. Price, C. Burgin, I. Catch and M. Cruise, Functional sensitivity and
recovery of thyroid-stimulating hormone, Clin Chem, 2001, 47, 2067.
86 L. M. Demers and C. A. Spencer, Laboratory medicine practice guidelines:
laboratory support for the diagnosis and monitoring of thyroid disease, Clin
Endocrinol (Oxf), 2003, 58, 138–40.
87 M. L. Rawlins and W. L. Roberts, Performance characteristics of six third-
generation assays for thyroid-stimulating hormone, Clin Chem, 2004, 50,
2338–44.
88 I. Sergi, M. J. Papandreou, G. Medri, C. Canonne, B. Verrier and C. Ronin,
Immunoreactive and bioactive isoforms of human thyrotropin, Endocrinology,
1991, 128, 3259–68.
89 S. Donadio, A. Pascual, J. H. Thijssen and C. Ronin, Feasibility study of new
calibrators for thyroid-stimulating hormone (TSH) immunoprocedures based
on remodeling of recombinant TSH to mimic glycoforms circulating in
patients with thyroid disorders, Clin Chem, 2006, 52, 286–97.
90 S. Donadio, A. Pascual, M. Dugas and C. Ronin, Standardization of TSH
immunoassays: production of new calibrators and harmonization of assays,
Immunoanalyse & Biologie Spećialiseé, 2006, 21, 79–90.
Carbohydr. Chem., 2012, 38, 94–123 | 119

91 European Medicines Agency, Pre-authorisation Evaluation of Medicines for
Human Use Committee for medicinal products for human use (CHMP)
Guideline on development, production, characterization and specifications for
monoclonal antibodies and related products BWP/157653 2008 (accessed
September 2, 2011 at http://www.ema.europa.eu/docs/en_GB/document_
library/Scientific_guideline/2009/09/WC500003074.pdf)
92 Center for Biologics Evaluation and Research (CBER) Center for Drug
Evaluation and Research (CDER) FDA Guidance concerning demonstration
of comparability of human biological products, including therapeutic bio-
technology-derived products 1996 (accessed September 2, 2011 at http://
www.fda.gov/Drugs/GuidanceComplianceRegulatoryInformation/Guidances/
ucm122879.htm)
93 International conference on harmonization of technical requirements for
registration of pharmaceuticals for human use. ICH harmonized Tripartite
Guideline. Specifications: Test procedures and acceptance for biotechnologi-
cal/biological Products Q6B 1999 (accessed September 2, 2011 at http://
www.ich.org/fileadmin/Public_Web_Site/ICH_Products/Guidelines/Quality/
Q6B/Step4/Q6B_Guideline.pdf)
94 K. N. Baker, M. H. Rendall, A. E. Hills, M. Hoare, R. B. Freedman and D. C.
James, Metabolic control of recombinant protein N-glycan processing in NS0
and CHO cells, Biotechnol Bioeng, 2001, 73, 188–202.
95 P. Hossler, S. F. Khattak and Z. J. Li, Optimal and consistent protein gly-
cosylation in mammalian cell culture, Glycobiology, 2009, 19, 936–49.
96 M. Gawlitzek, M. Estacio, T. Fürch and R. Kiss, Identification of cell culture
conditions to control N-glycosylation site-occupancy of recombinant glyco-
proteins expressed in CHO cells, Biotechnol Bioeng, 2009, 103, 1164–75.
97 D. C. Andersen, T. Bridges, M. Gawlitzek and C. Hoy, Multiple cell culture
factors can affect the glycosylation of Asn-184 in CHO-produced tissue-type
plasminogen activator, Biotechnol Bioeng, 2000, 70, 25–31.
98 D. C. Wong, N. S. Wong, J. S. Goh, L. M. May and M. G. Yap, Profiling of
N-glycosylation gene expression in CHO cell fed-batch cultures, Biotechnol
Bioeng, 2010, 15, 516–28.
99 D. C. Andersen, Cell culture effects on the glycosylation of therapeutic pro-
teins. Bioprocess International IBC Life Sciences, 2004.
100 X. Gu and D. I. Wang, Improvement of interferon-gamma sialylation in
Chinese hamster ovary cell culture by feeding of N-acetylmannosamine,
Biotechnol Bioeng, 1998, 58, 642–8.
101 Q. Zhou, S. Shankara, A. Roy, H. Qiu, S. Estes, A. McVie-Wylie, K.
Culm-Merdek, A. Park, C. Pan and T. Edmunds, Development of a simple
and rapid method for producing non-fucosylatedoligomannose containing
antibodies with increased effector function, Biotechnol Bioeng, 2008, 99, 652–65.
102 P. Taupin, Modulation of GDP-fucose level for generating proteins with
reduced rate of fucosylation (WO2010141855), Expert Opin Ther Pat, 2011,
21, 1497–9.
103 N. Munzert, E. Mthing, J. Büntemeyer, H and J. Lehmann, Sialidase activity
in culture fluid of Chinese hamster ovary cells during batch culture and its
effects on recombinant human antithrombin III integrity, Biotechnol Prog,
1996, 12, 559–63.
104 X. Gu, B. J. Harmon and D. I. Wang, Site- and branch-specific sialylation of
recombinant human interferon-gamma in Chinese hamster ovary cell culture,
Biotechnol Bioeng, 1997, 55, 390–8.
105 M. J. Gramer, C. F. Goochee, V. Y. Chock, D. T. Brousseau and M. B.
Sliwkowski, Removal of sialic acid from a glycoprotein in CHO cell culture
120 | Carbohydr. Chem., 2012, 38, 94–123

supernatant by action of an extracellular CHO cell sialidase, Biotechnology
(N Y), 1995, 13, 692–8.
106 M. Sarkar, E. Hull, Y. Nishikawa, R. J. Simpson, R. L. Moritz, R. Dunn and
H. Schachter, Molecular cloning and expression of cDNA encoding the
enzyme that controls conversion of high-mannose to hybrid and complex
N-glycans: UDP-N-acetylglucosamine: alpha-3-D-mannoside beta-1,2-N-
acetylglucosaminyltransferase I, Proc Natl Acad Sci U S A, 1991, 88, 234–8.
107 J. C. Paulson and K. J. Colley, Glycosyltransferases. Structure, localization, and
control of cell type-specific glycosylation, J Biol Chem, 1989, 264, 17615–8.
108 Y. Zhu, X. Li, A. Mcvie-Wylie, C. Jiang, B. L. Thurberg, N. Raben, R. J.
Mattaliano and S. H. Cheng, Carbohydrate-remodelled acid a-glucosidase
with higher affinity for the cation-independent mannose 6-phosphate receptor
demonstrates improved delivery to muscles of Pompe mice, Biochem J, 2005,
389, 619–628.
109 T. S. Raju, J. B. Briggs, S. M. Chamow, M. E. Winkler and A. J. Jones,
Glycoengineering of therapeutic glycoproteins: in vitro galactosylation and
sialylation of glycoproteins with terminal N-acetylglucosamine and galactose
residues, Biochemistry, 2001, 40, 8868–76.
110 A. Hayashi, T. Chiba, H. Hayashi, S. Sasayama, T. Ishiguro and K. Onozaki,
Synthesis of glycosylated human tumor necrosis factor alpha coupled with
N-acetylneuraminic acid, Cancer Immunol Immunother, 2007, 56, 545–53.
111 E. Grabenhorst, P. Schlenke, S. Pohl, M. Nimtz and H. S. Conradt, Genetic
engineering of recombinant glycoproteins and the glycosylation pathway in
mammalian host cells, Glycoconj J, 1999, 16, 81–97.
112 H. Narimatsu, Human glycogene cloning: focus on beta 3-glycosyltransferase
and beta 4-glycosyltransferase families, Curr Opin Struct Biol, 2006, 16,
567–75.
113 J. R. Desjarlais and G. A. Lazar, Modulation of antibody effector function,
Exp Cell Res, 2011, 317, 1278–85.
114 C. Ferrara, P. Brünker, T. Suter, S. Moser, U. Püntener and P. Umaña,
Modulation of therapeutic antibody effector functions by glycosylation
engineering: influence of Golgi enzyme localization domain and co-expression
of heterologous beta1, 4-N-acetylglucosaminyltransferase III and Golgi
alpha-mannosidase II, Biotechnol Bioeng, 2006, 5, 851–61.
115 M. S. Czuczman and S. A. Gregory, The future of CD20 monoclonal antibody
therapy in B-cell malignancies, Leuk Lymphoma, 2010, 51, 983–94.
116 N. Yamane-Ohnuki and M. Satoh, Production of therapeutic antibodies with
controlled fucosylation, MAbs, 2009, 1, 230–6.
117 N. Yamane-Ohnuki, S. Kinoshita, M. Inoue-Urakubo, M. Kusunoki, S. Iida,
R. Nakano, M. Wakitani, R. Niwa, M. Sakurada, K. Uchida, K. Shitara and
M. Satoh, Establishment of FUT8 knockout Chinese hamster ovary cells: an
ideal host cell line for producing completely defucosylated antibodies with
enhanced antibody-dependent cellular cytotoxicity, Biotechnol Bioeng, 2004,
87, 614–22.
118 T. T. Junttila, K. Parsons, C. Olsson, Y. Lu, Y. Xin, J. Theriault, L. Crocker,
O. Pabonan, T. Baginski, G. Meng, K. Totpal, R. F. Kelley and M. X.
Sliwkowski, Superior in vivo efficacy of afucosylated trastuzumab in the
treatment of HER2-amplified breast cancer, Cancer Res, 2010, 1, 4481–9.
119 Y. Kanda, H. Imai-Nishiya, R. Kuni-Kamochi, K. Mori, M. Inoue, K.
Kitajima-Miyama, A. Okazaki, S. Iida, K. Shitara and M. Satoh, Establish-
ment of a GDP-mannose 4,6-dehydratase (GMD) knockout host cell line:
a new strategy for generating completely non-fucosylated recombinant
therapeutics, J Biotechnol, 2007, 30, 300–10.
Carbohydr. Chem., 2012, 38, 94–123 | 121

120 H. Imai-Nishiya, K. Mori, M. Inoue, M. Wakitani, S. Iida, K. Shitara and M.
Satoh, Double knockdown of alpha1,6-fucosyltransferase (FUT8) and GDP-
mannose 4,6-dehydratase (GMD) in antibody-producing cells: a new strategy
for generating fully non-fucosylated therapeutic antibodies with enhanced
ADC, BMC Biotechnol, 2007, 7, 84.
121 H. H. Von Horsten, C. Ogorek, V. Blanchard, C. Demmler, C. Giese, K.
Winkler, M. Kaup, M. Berger, I. Jordan and V. Sandig, Production of non-
fucosylated antibodies by co-expression of heterologous GDP-6-deoxy-D-
lyxo-4-hexulose reductase, Glycobiology, 2010, 20, 1607–18.
122 T. Ishiguro, S. Kawai, K. Habu, M. Sugimoto, H. Shiraiwa, S. Iijima, S.
Ozaki, T. Matsumoto and H. Yamada-Okabe, A defucosylated anti-CD317
antibody exhibited enhanced antibody-dependent cellular cytotoxicity against
primary myeloma cells in the presence of effectors from patients, Cancer Sci,
2010, 101, 2227–33.
123 M. F. Naso, S. H. Tam, B. J. Scallon and T. S. Raju, Engineering host cell
lines to reduce terminal sialylation of secreted antibodies, MAbs, 2010, 2,
519–27.
124 S. Wildt and T. U. Gerngross, The humanization of N-glycosylation pathways
in yeast, Nat Rev Microbio, 2005, 3, 119–28.
125 J. Ferrari, J. Gunson, J. Lofgren, L. Krummen and T. G. Warner, Chinese
hamster ovary cells with constitutively expressed sialidase antisense RNA
produce recombinant DNase in batch culture with increased sialic acid, Bio-
technol Bioeng, 1998, 5, 589–95.
126 F. A. Ngantung, P. G. Miller, F. R. Brushett, G. L. Tang and D. I. Wang,
RNA interference of sialidase improves glycoprotein sialic acid content
consistency, Biotechnol Bioeng, 2006, 95, 106–19.
127 M. Zhang, K. Koskie, J. S. Ross, K. J. Kayser and M. V. Caple, Enhancing
glycoprotein sialylation by targeted gene silencing in mammalian cells, Bio-
technol Bioeng, 2010, 105, 1094–105.
128 D. Jones, N. Kroos, R. Anema, B. van Montfort, A. Vooys, S. van der,
Kraats, E. van der, Helm, S. Smits, J. Schouten, K. Brouwer, F. Lagerwerf, P.
van Berkel, D. J. Opstelten, T. Logtenberg and A. Bout, Biotechnol. Prog.,
2003, 19(1), 163–168.
129 K. Bork, R. Horstkorte and W. Weidemann, Increasing the sialylation of
therapeutic glycoproteins: the potential of the sialic acid biosynthetic pathway,
J Pharm Sci, 2009, 98, 3499–508.
130 K. Fukuta, R. Abe, T. Yokomatsu, N. Kono, M. Asanagi, F. Omae, M. T.
Minowa, M. Takeuchi and T. Makino, Remodeling of sugar chain structures
of human interferon-gamma, Glycobiology, 2000, 10, 421–30.
131 K. Fukuta, T. Yokomatsu, R. Abe, M. Asanagi and T. Makino, Genetic
engineering of CHO cells producing human interferon-gamma by transfection
of sialyltransferases, Glycoconj J, 2000, 17, 895–904.
132 S. Weikert, D. Papac, J. Briggs, D. Cowfer, S. Tom, M. Gawlitzek, J. Lofgren,
S. Mehta, V. Chisholm, N. Modi, S. Eppler, K. Carroll, S. Chamow, D. Peers,
P. Berman and L. Krummen, Engineering Chinese hamster ovary cells to
maximize sialic acid content of recombinant glycoproteins, Nat Biotechnol,
1999, 17, 1116–21.
133 Y. T. Jeong, O. Choi, Y. D. Son, S. Y. Park and J. H. Kim, Enhanced
sialylation of recombinant erythropoietin in genetically engineered Chinese-
hamster ovary cells, Biotechnol Appl Biochem, 2009, 52, 283–91.
134 O. T. Keppler, S. Hinderlich, J. Langner, R. Schwartz-Albiez, W. Reutter and
M. Pawlita, UDP-GlcNAc 2-epimerase: a regulator of cell surface sialylation,
Science, 1999, 84, 1372–6.
122 | Carbohydr. Chem., 2012, 38, 94–123

135 K. Bork, W. Reutter, W. Weidemann and R. Horstkorte, Enhanced sialylation
of EPO by overexpression of UDP-GlcNAc 2-epimerase/ManAc kinase
containing a sialuria mutation in CHO cells, FEBS Lett, 2007, 581, 4195–8.
136 Y. D. Son, Y. T. Jeong, S. Y. Park and J. H. Kim, Enhanced sialylation
of recombinant human erythropoietin in Chinese hamster ovary cells by
combinatorial engineering of selected genes, Glycobiology, 2011, 21, 1019–28.
137 E. U. Lee, J. Roth and J. C. Paulson, Alteration of terminal glycosylation
sequences on N-linked oligosaccharides of Chinese hamster ovary cells by
expression of beta-galactoside alpha 2,6-sialyltransferase, J Biol Chem, 1989,
264, 13848–55.
138 L. Monaco, A. Marc, A. Eon-Duval, G. Acerbis, G. Distefano, D. Lamotte,
J. M. Engasser, M. Soria and N. Jenkins, Genetic engineering of a2,6-sialyl-
transferase in recombinant ClIO cells and its effects on the sialylation of
recombinant interferon gamma, Cytotechnology, 1996, 22, 197–203.
139 X. Zhang, S. H. Lok and O. L. Kon, Stable expression of human alpha-2,6-
sialyltransferase in Chinese hamster ovary cells: functional consequences for
human erythropoietin expression and bioactivity, Biochim Biophys Acta, 1998,
1425, 441–52.
140 A. Bragonzi, G. Distefano, L. D. Buckberry, G. Acerbis, C. Foglieni, D.
Lamotte, G. Campi, A. Marc, M. R. Soria, N. Jenkins and L. Monaco, A new
Chinese hamster ovary cell line expressing alpha2,6-sialyltransferase used as
universal host for the production of human-like sialylated recombinant
glycoproteins, Biochim Biophys Acta, 2000, 1474, 273–82.
141 R. Jassal, N. Jenkins, J. Charlwood, P. Camilleri, R. Jefferis and J. Lund,
Sialylation of human IgG-Fc carbohydrate by transfected rat alpha2,6-
sialyltransferase, Biochem Biophys Res Commun, 2001, 286, 243–9.
142 P. Legaigneur, C. Breton, A. El Battari, J. C. Guillemot, C. Auge, M.
Malissard, E. G. Berger and C. Ronin, Exploring the acceptor substrate
recognition of the human beta-galactoside alpha 2,6-sialyltransferase, J Biol
Chem, 2001, 276, 21608–17.
143 S. Donadio, C. Dubois, G. Fichant, L. Roybon, J. C. Guillemot, C. Breton
and C. Ronin, Recognition of cell surface acceptors by two human alpha-2,6-
sialyltransferases produced in CHO cells, Biochimie, 2003, 85, 311–21.
144 Y. Liu, A. Nguyen, R. L. Wolfert and S. Zhuo, Enhancing the secretion of
recombinant proteins by engineering N-glycosylation sites, Biotechnol Prog,
2009, 25, 1468–75.
Carbohydr. Chem., 2012, 38, 94–123 | 123

Congenital Disorders of Glycosylation
(CDG): from glycoproteins to patient care
Vanessa Ferreira,*a Paz Brionesb and Maria-Antonia Vilasecac
DOI: 10.1039/9781849734769-00124
1 Introduction
Carbohydrates are at the forefront of Congenital Disorders of Glycosyla-
tion (CDG), a rapidly growing family of genetic diseases with more than
50 members identified at the molecular and biochemical level since the first
clinical description in 1980. They are due to defects in the synthesis of
glycans and in their attachment to proteins and lipids.
Glycosylation is one of the most complex post-translational modifications
and more than a half of human proteins are glycosylated.1 A disruption of the
glycosylation machinery can lead to a broad range of clinical phenotypes and
affect nearly every organ, but particularly the nervous system as most of these
disorders have neurological involvement. Although most of the deficiencies
observed in CDG patients are only partial, the severity of the clinical man-
ifestations underlines the relevance of glycosylation. Thus, advances in the
development of diagnostic tools and research contribute to a better under-
standing of the causes and mechanisms of CDG, improving (early) diagnosis,
increasing the knowledge and contributing to a better quality of life and
outcome of children and adults affected by these pathologies.
This chapter will cover developments of the past few years, highlighting
several methods and experimental systems that identified them, advantages
and limitations of biological model systems, and possible new treatments.
Finally, the role of CDG patients in the ‘‘glycobiomedical’’ field will be
briefly commented.
2 Basic principles of protein glycosylation

Posttranslational modification is a main step in the evolution of proteins
towards functional diversity. It extends the range of possible functions of a
protein through the introduction of new chemical groups that may, for
example, alter the hydrophobicity in membrane proteins or activate or
inactivate enzymes. Glycosylation is the most important and complex form of
posttranslational modification. Approximately 250 to 500 genes representing
B1–2% of the total number of human genes are involved in this process.2
Research over the last decade has conclusively shown that protein-
glycosylation is a vital process which plays a critical role in a large number
of biochemical and cellular processes including nascent chain folding,
cellular targeting, protease resistance, immunogenicity, signal transduction,
a
Vanessa Ferreira, Portuguese Association for CDG and other Rare Metabolic Diseases
(APCDG-DMR), Rua Manuel da Fonseca n1 46, 2820-381 Charneca da Caparica, Portugal.
E-mail: sindromecdg@gmail.com
b
CSIC. IBC, Sección de Errores Congénitos del Metabolismo, Hospital Clıńic. Servicio de
Bioquı´mica y Genética molecular. Mejıá Lequerica s/n. Edificio Helios III, planta baja. 08028
Barcelona, Spain.
c
Guia Metabólica, Passeig de Sant Joan de Deú n1 2. 08950. Esplugues de Llobregat, Spain.
124 | Carbohydr. Chem., 2012, 38, 124–155

pharmacokinetics, and protein secretion. The glycan chains regulate enzy-
matic activity, confer enhanced stability and solubility to secreted proteins,
and affect the functionality of immune proteins. However, the exact role of
glycans in a multitude of glycoproteins remains unclear. Hypoglycosylation
of glycoconjugates functioning as hormones, enzymes or transporters, leads
to their impaired bioavailability, decreased activity and/or rapid degrada-
tion. Most soluble and membrane-bound proteins are glycoproteins, in
contrast with the cytosolic proteins.
Glycosylation takes place both in the ER and Golgi. N-glycosylation
takes place cotranslationally in the ER and further processing occurs in the
Golgi. Whereas in the N-glycosylation carbohydrates are added as a large
preformed oligosaccharide from a carrier, named dolichol, always to
asparagine (Asn) as soon as it appears inside the ER, the O-glycosylation
occurs posttranslationally and is generally initiated in the cis-Golgi. In this
case, groups of the amino acids serine (Ser) or threonine (Thr) are the
acceptor sites of the protein.
2.1 N-glycosylation of proteins

N-glycosylation is the most prevalent modification of membrane and
secretory glycoproteins in eukaryotes. It is a three part process:
A) Assembly of the lipid-linked oligosaccharide (LLO).
B) En-bloc transfer of the LLO to the protein by the oligosaccharyl-
transferase (OST).
C) Remodeling of the oligosaccharide (glycan) once attached to the
protein by removal and addition of monosaccharides.
A) Assembly of the lipid-linked oligosaccharide (LLO): the dolichol cycle.
The biosynthesis of the oligosaccharide starts at the cytoplasmic site of
the ER membrane in a step-by-step process on the carrier lipid dolichol
phosphate (Dol-P). Dolichol is an isoprenoid compound synthesized by the
same metabolic route as cholesterol and it is phosphorylated by a kinase that
uses cytidine triphosphate (CTP) to form Dol-P. The availability of Dol-P at
the cytoplasmic site of the ER membrane is a rate limiting factor in the
dolichol cycle and N-glycosylation.3 Dol-P is the structure upon which the
carbohydrate moieties of N-linked glycoproteins are built. The dolichol cycle
has been the focus of intensive research, and its strong conservation among
higher systems has been an advantage in the elucidation of novel CDG.
Different nucleotide-activated sugar donors synthesized from mono-
saccharides are required to form the oligosaccharide precursor. In the cytosol,
the donors are uridine-diphospho-N-acetylglucosamine (UDP-GlcNAc) as
N-acetylglucosamine (GlcNAc) donor and guanosinediphosphate-mannose
(GDP-Man) as mannose (Man) donor. GDP-Man is generated from man-
nose-6-phosphate (Man-6-P), which is derived from the glucose, fructose, and
mannose metabolism. Man-6-P can be formed either from fructose 6-phos-
phate, via phosphomannose isomerase (PMI) and the glucose pathway, or
from free intracellular mannose phosphorylation by hexokinase (Hk) indi-
cating that the synthesis of LLO is also limited by the pool of free mannose in
the cell. Mannose-6-phosphate (Man-6-P) is then isomerized by the action of
phosphomannomutase 2 (PMM2) into mannose-1-phosphate (Man-1-P).
Carbohydr. Chem., 2012, 38, 124–155 | 125

Glucose Fructose Mannose
Blood vessel
HK HK HK Cytosol
GDP-M synthase
-6-P -6-P -6-P -1-P -GDP
PMI PMM2
-1-P
-UDP -UDP
PMI: Phosphomannose isomerase

N-acetylglucosamine PMM2: Phosphomannomutase 2
Lipid-linked
Oligosaccharide (LLO) Mannose
HK: Hexokinase
CDG defect
GDP-M synthase: Guanosine-
monophosphate synthase
Fig. 1 Metabolic pathway showing the sources and utilization of mannose in N-glycosylation is
shown. Glucose, fructose and mannose are three of the major forms of sugar in humans from
which all other monosaccharides can be synthesized. Through the action of several enzymes
these sugars are converted into various nucleotide sugar donors used to form complex glycans
in humans. The synthesis of nucleotide activated sugars is under tight regulatory control. Loss
of this control due to enzyme deficiency can significantly impair glycosylation with potential
serious clinical manifestations as it causes shortage in GDP-Man and Dol-P-Man, both needed
for the proper synthesis of lipid-linked oligosaccharide (LLO). For example, phosphomanno-
mutase (PMM) or phosphomannose isomerase (PMI) deficiency causes PMM2-CDG (CDG
Ia) and PMI-CDG (CDG Ib) respectively. Most mannose that is taken up is converted to
fructose-6-P via the combined actions of hexokinase (HK) and phosphomannose isomerase
(PMI). The mannose-6-phosphate (Man-6-P) is isomerised by the action of phosphomanno-
mutase 2 (PMM2) into mannose-1-phosphate (Man-1-P). Note the important position of the
PMI and PMM2 enzymes and the entry of mannose into the pathway to understand why
treatment is available for PMI-CDG (CDG-Ib).
The latter is subsequently converted to GDP-Man by the action of guano-

sine-monophosphate transferase4 (Fig. 1).
The oligosaccharide precursor assembly is initiated by the transfer of two
N-acetylglucosamine residues to Dol-P from UDP-GlcNAc. The product is
then extended by single and sequential mannose additions through the
action of different mannosyltransferases using GDP-Man as the donor
substrate. Once the two N-acetylglucosamines (GlcNAc) and five mannoses
have been added, the Dol-PP-glycan flips from the cytosol to the ER lumen
by the action of a flippase. In the ER lumen, mannosyl- and glucosyl-
transferases are responsible for further elongation of the oligosaccharide
precursor using Dol-P linked sugar donors (Dol-P-Man and Dol-P-Glc).
B) ‘‘En-bloc’’ transfer of the oligosaccharide to the protein by oligo-
saccharyltransferase (OST). The final 14 monosaccharide glycan
(Glc3Man9GlcNAc2), is subsequently transferred to an asparagine residue
contained in a N-X-S(T) sequon of the polypeptide substrate, where X can be
any amino acid except proline (glycosylation consensus site),5 and releasing
Dol-PP. This is catalyzed by the OST complex, a hetero-oligomeric membrane
126 | Carbohydr. Chem., 2012, 38, 124–155

protein that in humans consists of seven transmembrane proteins: ribophorin I,
ribophorin II, OST48, OST4, STT3A/ STT3B, N33/IAP, and DAD1. Studies in
yeast showed that OST subunits N33/IAP and OST4 have accessory functions,
while the other subunits are proposed to participate directly in N-glycosylation
and are essential for yeast viability. Mammalian and other eukaryotic organ-
isms express several OST isoforms.6 For example, the human variants STT3A
and STT3B are co-expressed in different tissues but differ in expression levels;
OST isoforms harbouring one or the other STT3 protein differ in catalytic
activity and substrate specificity. By expressing different OST isoforms,
organisms may respond to alterations of the glycoprotein flux through the
secretory pathway and this may contribute to tissue-specific differences in the
glycan repertoire. In vitro studies have shown that OST has a high affinity for
the fully assembled oligosaccharide, even if its abundance is lower, whereas
shorter intermediates are transferred inefficiently. For this reason, many dif-
ferent defects leading to incomplete LLO will result in a lack of the complete
oligosaccharide on many different glycoproteins at some glycosylation sites.
This will affect their functionality and will result in the broad spectrum of
symptoms that characterize CDG-I patients. In contrast, recent descriptions
of highly specific clinical pictures in patients with mutations in the subunits of
OST suggest that OST deficiencies lead to more restricted phenotypes.7
C) Remodeling of the oligosaccharide by removal and addition of mono-

saccharides. Subsequent processing of the core sugar (Glc3Man9GlcNAc2)
attached to protein starts in the ER by the trimming of the three glucose
residues and one mannose. This process is crucial for the quality control of
protein folding. High-mannose oligosaccharides interact with the chaper-
ones calreticulin and calnexin which serve as sensors by binding to mono-
glucosylated proteins, facilitating their folding and assembly. Only
glycoproteins that have been correctly folded get packaged into ER-to-
Golgi transport vesicles. In the cis Golgi, the processing continues with
further trimming of mannose residues and addition of one GlcNAc to form
branched structures. Addition of N-acetylglucosamine (GlcNAc), galactose
(Gal), fucose (Fuc) and sialic acid (N-acetyl-neuraminic acid, NANA)
residues in the medial-and trans-Golgi will result in the mature glycopro-
teins to be transported to their destination in the cell, in the cell membrane
and outside the cell. To this purpose nucleotide sugar transporters are
carried into the Golgi8 (Fig. 2).
2.2 O-glycosylation of proteins

In contrast with N-glycosylation, the O-glycosylation processes do not
follow a unique model, and therefore, they produce an immense multitude
of more or less branched chemical structures. Seven different types are
found in humans that are classified according to the first sugar residue
linked to the hydroxyl group of a serine, threonine or hydroxylysine of the
protein. O-glycosylation has no processing and thus only consists of
assembly. Contrary to N-glycosylation, this assembly mainly occurs in the
Golgi. Examples of important O-glycans are O-N-acetylglucosaminyl-
glycans (mucin-type glycans), O-xylosylglycans (glycosaminylglycans),
O-mannosylglycans, and O-fucosylglycans (Table 1).
Carbohydr. Chem., 2012, 38, 124–155 | 127

CDG-I CDG-II
Oligosaccharyl-
transferase
Dolichol
Lipid-linked
Nascent Glycan Mature
oligosaccharide
glycoprotein remodeling glycoproteins
(LLO) assembly
Monosaccharides
ER Quality
control
CYTOSOL ER LUMEN GOLGI LUMEN
Fig. 2 Overview of the N-glycosylation process and its cellular context. Congenital Disorders of
Glycosylation can be classified as type I and type II, depending on the nature and localization
of the biochemical defect. Activated sugars are necessary for lipid-linked oligosaccharide
assembly in the ER. This is followed by processing of N-linked oligosaccharides, which involves
deletions, additions, and rearrangements of sugars starting in the ER and finishing in the Golgi.
Upon completion of glycan remodeling, the mature glycoproteins are transported to their
destinations. Abnormally glycosylated proteins due to CDGs are removed through specific ER
quality controls (Figure adapted from Struewe et al., 2009).
Table 1 Different types of O-linked glycans in humans (modified from Wopereis et al.9).
Type of O-linked glycan Glycoprotein
Mucin–type (O-linked GalNAc) Secreted and plasma membrane

GAGa Proteoglycans
O-linked GlcNAc Nuclear and cytoplasmic
O-linked Gal Collagens
O-linked Man a-dystroglycan
O-linked Glc EGF protein domains
O-linked Fuc EGF protein domains; TSR repeats
a
glycosaminoglycans.
2.3 N- and O-glycosylation

Part of the glycosylation machinery (some glycosyltransferases, sugar-
nucleotide transporters, etc. . .) is common to N- and O-glycosylation.
In addition, recent studies showed that some proteins are essential
for establishing and/or maintaining the structure and function of the
Golgi, such as the Conserved Oligomeric Golgi (COG) complex, which is an
eight-subunit (Cog 1-8) peripheral Golgi protein involved in retrograde
vesicular Golgi trafficking.10 ATPase 6 is also essential for proper main-
tenance of the Golgi.11
3 Congenital disorders of glycosylation (CDG)

3.1 CDG classification
CDG are due to defects in the synthesis of the glycan moiety of glyco-
proteins and glycolipids, and in their attachment to proteins and lipids.
128 | Carbohydr. Chem., 2012, 38, 124–155

There are two main groups of N-glycosylation defects. CDG-I comprises
defects in the assembly of the dolichol pyrophosphate oligosaccharide and
in the transfer of the oligosaccharide from dolichol to the nascent poly-
peptides in the ER. CDG-II consists of defects in the processing of the
protein-linked glycans in the ER and the Golgi compartment (Fig. 3). O-
glycosylation, defects have been identified in the synthesis of O-xylosylglycans,
O-N-acetylgalactosaminylglycans, O-xylosyl/N-acetylgalactosaminylglycans,
O-mannosylglycans and of O-fucosylglycans (Table 1).
Fig. 3 Overview of glycoprotein biosynthesis and known CDG types. Dolichol is a long chain
lipid strongly bound to the ER membrane. On the cytosolic side of the ER membrane, GlcNAc
from UDP- GlcNAc, and mannose from GDP-Man are enzymatically linked to dolichol-PP
with concomitant cleavage of the nucleotide moiety. As mentioned in the text, the preassembled
carbohydrate structure linked to dolichol-PP is commonly referred to as the lipid-linked oli-
gosaccharide (LLO). When the LLO structure is Man5GlcNAc2–PP-dolichol, the glycan is
flipped into the lumen of the ER by a ‘‘flippase’’ and the synthesis is continued by the action of
mannosyl and glucosyltransferases until the final structure Glc3Man9GlcNAc2 –PP-dolichol.
This completed oligosaccharide is transfered ‘‘en bloc’’ to the nascent protein by a oligo-
saccharyltransferase (OST). The processing of its composition proceeds in the ER and later in
the Golgi apparatus. The action of various glucosidases, mannosidases, GlcNAc-transferases,
galactosyltransferases, sialyltransferases, and eventually fucosyltransferases lead to a final-
carbohydrate core. All CDG defects identified so far are symbolized by a black bar and the
corresponding gene in italics (Figure adapted from Körner et al., 2011).
Carbohydr. Chem., 2012, 38, 124–155 | 129

There is also a growing group of CDG with combined N- and O-
glycosylation defects. A number of these are defects in proteins with a
broader range of functions besides glycosylation. The first description of
this type of combined defects was on COG-CDG. These defects illustrate
the importance of the correct environment for glycosylation. COG defects
cause alterations in the control of the intracellular trafficking and/or
stability of a variety of Golgi-associated glycosylation enzymes.
The subsequent description of dolichol phosphate synthesis defects,
ATP6V0A2-CDG, and SEC23B-CDG has led to the distinction of two
groups of CDG: the defects in the glycosylation machinery and the group of
defects in proteins that are necessary for creating the environment that
permits the proper functioning of this machinery.12
An overview of glycoprotein biosynthesis and known CDG types is
presented in Fig. 3. The disorders are showed according to the gene defect.
A growing number of patients with as yet unclassified CDG (CDG-X)
might have similar ‘‘environmental’’ problems, indicating that in the near
future other proteins involved in Golgi structure and function may be
described and identified as responsible for new CDGs.13–19
Of the about 50 reported CDG, 17 affect N-glycosylation, 11 affect O-
glycosylation, 3 involve defects in glycosphingolipid and glycosylphos-
phatidylinositol anchor glycosylation and 19 are caused by defects in
multiple glycosylation pathways and other pathways.
The present knowledge of CDG types identified and the approximate
number of affected patients are listed in Table 2. The novel nomenclature
(introduced in 2009) is used along with the traditional nomenclature pre-
sented in parenthesis (for example, Ia for CDG-Ia).20–22
Approximately, 1000 patients with N-glycosylation defects have been
published, of which PMM2-CDG (CDG-Ia) is by far the most common
worldwide. Most CDG subtypes have been described in only a few indivi-
duals. Therefore, the phenotypic spectrum of these diseases is incompletely
known. Since there are a large number of steps involved in glycosylation
and related pathways, it is expected that this group of disorders will expand
rapidly in the near future.
3.2 Clinical manifestations of CDG

Since glycosylation has a general importance, it is not surprising, that CDG
often result in severe and multisystem disease. In fact, CDG cuts across
nearly all medical specialties.
The clinical picture is variable but nearly always comprises dysmorphism,
neurological involvement (intellectual impairment, ataxia, seizures, retino-
pathy), and other organ involvement (hepatic dysfunction, coagulopathies,
failure to thrive, cardiomyopathy and pericardial effusion, hydrops fetalis,
endocrine abnormalities, renal dysfunction, and skeletal defects). CDG
should thus be considered in every patient with an unexplained syndrome.
Limited awareness of CDG contributes to underdiagnosis of these dis-
orders. Details about the clinical spectrum of CDG are summarized in
Table 2 and are briefly mentioned here because they have been described in
numerous excellent reviews (see references in Table 2).
130 | Carbohydr. Chem., 2012, 38, 124–155

Table 2 Clinical manifestations of human genetic disorders of glycosylation (Adapted from67–71).
Defects of protein N-glycosylation

CDG type
Number of reported Screening technique/alternative
patients Protein Overview of key cinical manifestations technique
PMM2-CDG (Ia) Phosphomannomutase Neurology: strabismus, ataxia, hypotonia, psychomotor retardation, Tf-IEF type 1
W80072 peripheral neuropathy, epilepsy, stroke-like episodes,
(olivoponto)cerebellar hypoplasia
Dysmorphism: inverted nipples, fat pads
Gastroenterology/hepatology: failure to thrive, recurrent vomiting, liver
involvement, pancreatitis
Growth retardation
Neonatology: hydrops fetalis, floppy baby
Haematology: thrombosis, increased platelet adhesiveness, clotting and
anticlotting factor abnormalities
Endocrinology: hypothyroidism, hypergonadotrophic hypogonadism,
hyperinsulinaemic hypoglycaemia
Orthopaedics: kyphoscoliosis, joint contractures, osteopenia
Cardiology: pericardial effusion, cardiomyopathy
Ophthalmology: retinitis pigmentosa
Multi-organ failure73–97
PMI-CDG (Ib) Phosphomannose isomerase Gastroenterology/hepatology: failure to thrive, protein-losing enteropathy, Tf-IEF type 1
W2572 life-threatening intestinal bleeding, prolonged diarrhoea, congenital
hepatic fibrosis, increased serum transaminases
Endocrinology: hypoglycaemia34–39,98–101
ALG6-CDG (Ic) Glucosyltransferase I Neurology: moderate to severe psychomotor retardation, seizures, Tf-IEF type 1
W3672 hypotonia, pseudotumor cerebri, epilepsy, strabismus
Gastroenterology/hepatology: feeding difficulties
Haematology: deep vein thrombosis
Endocrinology: hyperandrogenism and virilization
Carbohydr. Chem., 2012, 38, 124–155 | 131

Others: limb deficiencies resembling brachydactyly type B102–112
Table 2 (Continued )
Defects of protein N-glycosylation

CDG type
Number of reported Screening technique/alternative
patients Protein Overview of key cinical manifestations technique
ALG3-CDG (Id) Mannosyltransferase VI Neurology: slowly progressive encephalopathy with microcephaly, severe Tf-IEF type 1
1172 psychomotor retardation and epileptic seizures
Gastroenterology/hepatology: failure to thrive
Orthopaedics: osteopenia113–116
ALG12-CDG (Ig) Mannosyltransferase VIII Neurology: microcephaly, psychomotor retardation Tf-IEF type 1
Gastroenterology/hepatology: intestinal malrotation with poor gastrointestinal
motility
Neonatology: short-limb skeletal dysplasia, polyhydramnios, prematurity, and
generalized edema
Endocrinology: small penis with hypospadias, hypoplastic scrotum, and non-
palpable testes
Dysmorphic features: included broad nose, thick ears, thin lips, micrognathia,
132 | Carbohydr. Chem., 2012, 38, 124–155

inverted nipples, ulnar deviation at the wrists, spatulate fingers, fifth finger
camptodactyly, nail hypoplasia, and talipes equinovarus
Cardiology: PDA, VSD, hypertrophic cardiomyopathy, and arrhythmias.
Orthopaedics: absent ossification of cervical vertebral bodies, pubic bones,
knee epiphyses, and tali
Others: sensorineural deafness, retinal detachment and blindness117–122
ALG8-CDG (Ih) Glucosyltransferase II Neurology: multifocal myoclonic seizures, optic nerve atrophy, cerebellar Tf-IEF type 1
972 hypoplasia
Gastroenterology/hepatology: diarrhoea progressing to protein-losing entero-
pathy with ascites, oedema
Haematology: bleeding
Cardiology: pericardial effusion123–128
ALG2-CDG (Ii) Mannosyltransferase II Neurology: retardation, seizures, hypomyelination Tf-IEF type 1
1129 Gastroenterology/hepatology: hepatomegaly,
Haematology: clotting factor abnormalities (similar to PMM2-CDG)
Others: coloboma of the iris129
DPAGT1-CDG (Ij) GlcNActransferase I In general similar to PMM2-CDG. Tf-IEF type 1
372 Neurology: mental retardation very severe, severe hypotonia, medically
intractable seizures, microcephaly
Ophtalmology: exotropia130
ALG1-CDG (Ik) Mannosyltransferase I severe multisystem disorder with recurrent nonimmune hydrops fetalis131–134 Tf-IEF type 1
772
ALG9-CDG (Il) Mannosyltransferase VII Neurology: psychomotor retardation, esotropia, axial hypotonia, epilepsy Tf-IEF type 1
3135–137 delayed myelination, decreased volume of cerebral hemispheres and cerebellum
Dysmorphism: inverted nipples
Gastroenterology/hepatology: failure to thrive, hepatosplenomegaly
Cardiology: pericardial effusion
Others: cystic renal disease, low levels of multiple serum proteins including
antithrombin III, factor XI, and cholesterol135–138
RFT1-CDG (In) RFT1 protein Neurology: Developmental retardation, poor to absent visual contact, sensor- Tf-IEF type 1
612 ineural deafness, epilepsy, microcephaly, cerebral atrophy and hypotonia
Gastroenterology/hepatology: feeding problems and failure to thrive
Others: dysmorphism, respiratory problems and venous thromboses139,140
ALG11-CDG (Ip) Mannosyltransferase III Multisystem disorder with muscular hypotonia, seizures, developmental Tf-IEF type 1
312 retardation141,142
MGAT2-CDG (IIa) GlcNActransferase II Growth retardation, mental retardation, facial dysmorphism23,143,144 Tf-IEF type 2.
469 Normal apoC-III IEF.
Further confirmation by mass
spectrometry
GCS1-CDG (IIb) Glucosidase I (GCS1) Neurology: hypotonia, epilepsy; Normal Tf- and apoC-III IEF.
1145 Dysmorphism; Hepatic involvement145 Urine oligosaccharide analysis
TUSC3-CDG Oligosaccharyltransferase Isolated intellectual disability146–150 Normal Tf-IEF and Apo C-III
TUSC3 subunit IEF.
MAGT1-CDG Magnesium transporter Isolated intellectual disability150 Genetic analysis
Carbohydr. Chem., 2012, 38, 124–155 | 133

441 protein 1
Defects of protein O-glycosylation

CDG type Protein Overview of key clinical manifestations
EXT1/EXT2-CDG Extosin-1,-2 Multiple cartilage-capped tumors located primarily at the long bones, often
(Multiple cartilaginous exostoses) a short stature and secondary orthopedic deformities151
b4GALT7-CDG Glucuronyltransferase/ Psychomotor retardation, short stature, osteopenia of all bones, defective
(Ehlers-Danlos syndrome) N-acetylglucosaminyl-transferaseb-1, deciduous teeth, delayed wound healing, hypotonia, progeroid aspect152
369 4-galactosyltransferase 7
GALNT3-CDG Polypeptide N-acetylgalactosaminyl- Ectopic calcifications around major joints153

(O-N-acetyl-galactosaminyl-glycan transferase 3
defect
(familial tumoral calcinosis)
POMGNT1-CDG O-mannosyltransferase 1 O-mannose Neurology, myology: congenital muscular dystrophy, myoclonic jerks, mental
(Muscle-eye-brain disease) b-1,2-N-acetylglucosaminyltransferase retardation, hydrocephalus
Ophthalmology: congenital glaucoma and myopia, retinal hypoplasia, pallor
134 | Carbohydr. Chem., 2012, 38, 124–155

of the optic disc154
SCDO3-CDG O-fucose-specific b-1,3-N- Multiple vertebral segmentation defects and rib anomalies
(Spondylocostal dysostosis type 3) acetylglucosaminyltransferase
B3GALTL-CDG Beta1,3-glucosyltransferase Neurology: psychomotor retardation
(Peter-plus Syndrome) Ophthalmology: malformations of the eyes that include corneal opaqueness and
iridocorneal adhesions (Peters’ anomaly)
Dysmorphism: short stature, short limbs; some have cleft lip/palate155,156
SLC35D1-CDG ER UDP-glucuronic acid/UDP-N- Skeletal dysplasia comprising mainly platyspondyly, extremely short long bones,
acetylgalactosamine and small ilia with snail-like appearance157
Defects of glycosphingolipid and GPI-anchor glycosylation

CDG type Protein Overview of key clinical manifestations
SIAT9-CDG GM3 synthase Neurology: psychomotor retardation, epilepsy

(Amish infantile epilepsy)
PIGM-CDG Phosphatidylinositol glycan anchor Neurology: epilepsy
biosynthesis, class M Haematology: splanchnic vein thrombosis12
PIGV-CDG Phosphatidylinositol glycan anchor Hyperphosphatasia mental retardation syndrome (Mabry syndrome)
biosynthesis, class V Neurology: hypotonia, epilepsy
Dysmorphism: hypertelorism, long palpebral fissures, broad nasal bridge and tip, downturned
corners of the mouth, thin upper lip12
Defects of multiple glycosylation and other pathways

CDG type Screening technique/alternative
Number of reported patients Protein Overview of key clinical manifestations technique
DPM1-CDG (Ie) Dol-P-Man synthase 1 Neurology: severe developmental delay, seizures Tf-IEF type 1
872 Dysmorphism158,159
MPDU1-CDG (If) Dol-P-Man utilization Neurology: psychomotor retardation, epilepsy Tf-IEF type 1
572 protein Gastroenterology/hepatology: failure to thrive,
Dermatology: dry skin and scaling with erythroderma,
Ophthalmology: impaired vision, night blindness160,161
DPM3–CDG (Io) Dol-P-Man synthase 3 Neurology: stroke-like episode Tf-IEF type 1
172 Cardiology: cardiomyopathy, elevated serum CK, mildly
elevated serum transaminases72,162–164
GNE-CDG (Hereditary Glucosamine (UDP- Myopathy in adults with sparing of quadriceps muscles165
inclusion body myopathy) N-acetyl)-2-epimerase/
N-acetylmannosamine
kinase
SLC35C1-CDG (IIc) GDP-fucose transporter Neurology: mental retardation Tf-IEF and apo C-III are
(leukocyte adhesion deficiency Dysmorphism with growth retardation normal.
type II syndrome-LAD II) Others: recurrent infection, persistent high Alternative assay: Bombay blood
leukocytosis,166–172 group, MS (serum N-glycan
profiling)
Carbohydr. Chem., 2012, 38, 124–155 | 135

Defects of multiple glycosylation and other pathways

CDG type Screening technique/alternative
Number of reported patients Protein Overview of key clinical manifestations technique
B4GALT1-CDG (IId) Galactosyltransferase Neurology: hydrocephalus, muscular hypotonia, transient TF-IEF type 2.
168 cholestatis ApoCIII is normal and MS is next
Haematology: clotting factor abnormalities step in diagnostic process
Others: myopathy173,174
SLC35A1-CDG (IIf) CMP-NeuAc transporter Macrothrombocytopenia TF-IEF and ApoCIII are normal.
Alternative assay: CD15s on
leukocytes
DK1-CDG (Im) Dolichol kinase Neurology: muscular hypotonia, nystagmus, TF-IEF type 1
469 Cardiology: dilated cardiomyopathy
136 | Carbohydr. Chem., 2012, 38, 124–155

Dermatology: hyperkeratosis
Others: pulmonary infections175
SRD5A3-CDG (Iq) Steroid5areductase3 Neurology: psychomotor retardation Tf-IEF type 1
1169 Ophthalmology: variable eye malformations, including optic
nerve hypoplasia, retinal coloboma, congenital cataract
and glaucoma113,114
Dermatology: eczema, ichthyosis176,177
COG7-CDG (IIe) Cog7 Neurology: progressive, severe microcephaly, hypotonia TF-IEF type 2.
769 Growth retardation ApoCIII is abnormal.
Dysmorphism: wrinkled skin MS is next step in diagnostic process.
Gastroenterology/hepatology: feeding problems by
gastrointestinal pseudo-obstruction, failure to thrive,
Cardiology: septal defects
Others: episodes of hyperthermia16,19,178,179
COG1-CDG (IIg) Cog1 Cerebrocostomandibular-like syndrome Tf-IEF type 2.
369 (COG complex subunit 1) ApoCIII is abnormal.
MS is next step in diagnostic process.
COG4-CDG (IIj) Cog4 Neurology: epilepsy, microcephaly, ataxia, brain MRI Tf-IEF type 2.
115 abnormalities ApoCIII is abnormal.
Haematology: clotting factor abnormalities MS is next step in diagnostic process.
Others: recurrent infections15,180
COG5-CDG Cog5 Neurology: mild psychomotor retardation with delayed language Tf-IEF type 2.
development 18 ApoCIII is abnormal.
MS is next step in diagnostic process.
COG6-CDG Cog6 Neurology: intracranial bleeding with vitamin K deficiency, Tf-IEF type 2.
secondary epilepsy ApoCIII is abnormal.
Gastroenterology: vomiting, cholestasis MS is next step in diagnostic process.
Early fatal outcome181
COG8-CDG (IIh) Cog8 Neurology: epilepsy, strabismus, oculomotor apraxia, hypotonia, Tf-IEF type 2.
microcephaly, contractures, peripheral neuropathy, ataxia, ApoCIII is abnormal.
brain MRI abnormalities MS is next step in diagnostic process.
Dysmorphism: finger/toe abnormalities
Gastroenterology: Feeding/intestinal problems, increased serum
transaminases
Haematology: clotting factor abnormalities14,17
ATP6V0A2-CDG Vesicular H(þ)-ATPase Neurology: psychomotor retardation TF-IEF type 2 and apoCIII is
(Cutis Laxa type 2) subunit a2 Dysmorphism: growth delay, joint anomalies, congenital abnormal
cutis laxa ameliorating with age11,182–184
SEC23B-CDG (Congenital dysery- Protein transport protein Dyserythropoietic anemia12
thropoietic anemia type II-CDAII Sec23B isoform 1
or HEMPAS)
Carbohydr. Chem., 2012, 38, 124–155 | 137

Early diagnosis of any genetic disease and thus also of CDG is important
for several reasons such as the timely management of clinical problems,
genetic counseling, and efficient treatment (PMI-CDG) or partial therapy
(SLC35C1-CDG and PIGM-CDG).
3.3 Diagnosis of CDG

3.3.1 Screening techniques. Isoelectrofocusing of serum transferrin Tf
(Tf-IEF) is the most widely used method to screen for N-glycosylation
defects associated with sialic acid deficiency. Tf is an abundant serum
protein. Most serum transferrin is tetrasialotransferrin because most
molecules of this protein carry two disialylated biantennary oligosaccharide
chains resulting in a total of four terminal sialic acids. In addition, some of
these oligosaccharide chains may be triantennary or even tetraantennary,
which explains the presence of penta and hexasialo forms. Any defect in the
assembly or processing of these glycans results in hyposialylation. Since
sialic acid has a negative charge, there is a cathodal shift in an electric field.
Transferrin isoforms are separated by IEF in agarose or polyacrylamide
gels containing a pH gradient obtained by addition of ampholytes/
ampholynes Serum samples are treated with Fe3 þ for transferrin iron
saturation. After separation, isoforms are immunofixed using anti-Tf anti-
bodies and stained with Coomassie Brilliant Blue. This procedure permits
clear identification of the altered isoforms. The altered profiles may be
classified in two types. Type 1 corresponds to hypoglycosylation due to
defects in the assembly of the glycan in the cytosol or the endoplasmic
reticulum and is a consequence of the lack of glycan branches linked to the
protein. In this type, there is an increase of asialo and disialo transferrin and
a decrease of the tetrasialo band. The type 2 pattern is due to a defect in the
procesing of the glycan structure linked to the protein and shows an increase
of the less sialylated bands, mainly the tri-, di- and mono- and asialylated
ones.23
Although widely used, IEF is a laborious and time consuming technique
neither suitable for automation nor for accurate quantification. Therefore
other more quantitative laboratory techniques such as capillary zone elec-
trophoresis (CZE) and high pressure liquid chromatography (HPLC) have
been implemented in some laboratories.
Several analytical procedures have been developed for the analysis of
carbohydrate-deficient transferrin (CDT). They are extensively used for
the follow-up of chronic alcohol abuse. However they indicate only the
increase of hyposialylated isoforms, without identifying the altered
isoforms.
3.3.2 Techniques for further investigation of CDG. After an altered

pattern has been detected, it is necessary to find the enzymatic deficiency
responsible for the defect. First of all, secondary CDG have to be excluded
such as galactosemia, hereditary fructose intolerance, chronic alcoholism,
sepsis, hepatopathy or immaturity. In these situations, the normal pattern is
restored after adequate treatment.
If the pattern is type 1, the diagnostic algorithm is clear (Fig. 4). Since
PMM2-CDG (previously CDG-Ia) is by far the most frequent CDG the
138 | Carbohydr. Chem., 2012, 38, 124–155

DIAGNOSTIC PATHWAY
Clinical suspicion
Obtain serum sample
Sialotransferrin IEF
Type 1 pattern
Normal pattern Type 2 pattern
CDG-I
CDG not excluded CDG-II
Obtain skin biopsy
(or leukocytes) In serum Obtain skin biopsy
Enzyme investigations
PMM (CDG-Ia) ApoCIII IEF Enzyme investigations
PMI (CDG-Ib) Other assays
Other enzymes and assays (LLO)
Glycan analysis, other assays
MUTATION STUDIES
Fig. 4 Algorithm of the diagnostic pathway commonly used from clinical suspicion to reach a
definitive diagnosis.
first step is to measure PMM activity in fibroblasts or leucocytes.20 If the

clinical picture suggests phosphomannose isomerase deficiency (PMI-CDG,
previously CDG-Ib), PMI activity should be analyzed in leucocytes with
urgency because this is a treatable disorder and it is important to introduce
mannose supplementation before development of a severe enteropathy.
After PMM and PMI deficiencies have been discarded, the following step
is the study of the lipid-linked oligosacharides (LLO) in fibroblasts cultured
with radioactive mannose. The accumulated abnormal glycan(s) will often
point to the precise altered step. This method allows identification of most
of the type I defects.24 In defects before the addition of mannoses to
GlcNAc2-PP-dolichol the LLO accumulation is normal or low.
For type 2 patterns, the algorithm is not so well defined (Fig. 4). After
excluding secondary causes, the following step is the study of the isoforms
of serum apoCIII by IEF. This protein is purely O-glycosylated protein. IEF
of this protein permits to exclude or confirm core 1 mucin-type O-glyco-
sylation defects. A type 2 transferrin pattern together with an altered apo
C-III pattern is seen in some CDG types.25
In the presence of type 2 pattern, it is also indicated to analyse the
abnormal structures of the protein-linked glycans by electrospray mass
spectrometry (ES-MS) or MALDI-TOF, as they are useful tools to suggest
the candidate step responsible for the altered glycan synthesis.
Other tools that have been used for the diagnosis of CDG-II are PNA lectin
staining which detects galactose residues of O-glycans not protected by a
terminal sialic acid residue. Treatment of fibroblasts with Brefeldin A (BFA) is
used to assess a possible trafficking defect in the Golgi. BFA is a selective
inhibitor of the GTP-exchange factor for ADP ribosylation factor 1 (Arf1) and
causes the redistribution of Golgi proteins to the ER via retrograde transport.
In COG patients, the BFA-induced redistribution is significantly delayed.
Carbohydr. Chem., 2012, 38, 124–155 | 139

3.3.3 Final CDG diagnosis. To establish a definite diagnosis of CDG it is
necessary to demonstrate the protein deficiency and/or the genetic muta-
tion(s). Enzymatic procedures are not available for all the enzymes involved
in CDG, but besides PMM and PMI some other enzymes are measurable. In
all cases, it is indispensable to achieve a precise diagnosis by detection of the
disease-causing mutation(s) allowing for a reliable genetic counseling.
As numerous genes are involved in the glycosylation pathways and many
CDG-X patients remain undiagnosed, new approaches should be used for
their study. At present, whole exome sequencing (targeted exome capture) is
being used as a strategy to selectively sequence the coding regions of the
human genome to identify functional variations associated with a pheno-
type or pathology. The exome sequencing is based on a first step of
enrichment of the targeted regions of the DNA (mainly exome) by hybri-
dization with probes followed by next generations sequencing. The obtained
data are treated and analyzed to try to pinpoint the genetic defect.
3.4 Genetics of CDG

The known CDG are transmitted in an autosomal recessive way except for
EXT1/EXT2-CDG (autosomal dominant) and MAGTI-CDG (X-linked).
The most common type of CDG is due to mutations in the PMM2 gene
(MIM#601785), located on chromosome 16p13. It encodes the PMM pro-
tein (EC 5.4.2.8, PMM 2). Up to date more than one hundred disease-
causing mutations have been described in this gene (HGMDs professional
release, http://www.hgmd.cf.ac.uk/ac/index.php). The most common
(nearly 85%) are missense changes; patients with this type of changes are
usually compound heterozygous for two different missense mutations.
PMM2-CDG patients show a strong genetic heterogeneity, although one
severe mutation, p.R141H, has been found particularly common in all
cohorts of patients analyzed.26–29 Specific missense changes have been
functionally analyzed to improve the understanding of the molecular
mechanisms of the disease.27,30–32
3.5 Treatment of CDG

3.5.1 Specific treatment. An efficient, specific treatment is available for
only one CDG, namely PMI-CDG (CDG-Ib). In this CDG, GDP-Man can
still be synthesized from free mannose entering the cell and directly phos-
phorylated to Man-6-P, overcoming the defect. Thus, the simple daily oral
mannose supplementation is an effective treatment for patients with this
potentially lethal disorder.33–39
Some patients with SLC35C1-CDG or leukocyte adhesion deficiency type
II respond to simple supplementation of oral fucose. Finally, in PIGM-
CDG that causes histone hypoacetylation, butyrate increases PIGM tran-
scription and its supplementation was able to control seizures in the one
reported patient.40
3.5.2 Symptomatic treatments and preventive measures. Table 3 sum-
marizes symptomatic treatments for CDG. In addition, occupational ther-
apy, physical therapy, and speech therapy are importat. Of note is that
before surgery the consult of a hematologist, familiar with CDG, is
indicated.
140 | Carbohydr. Chem., 2012, 38, 124–155

Table 3 Symptomatic management for CDGs without specific treatment.
Clinical feature
Failure to thrive Consider nasogastric tube or gastrostomy tube feeding41

Strabismus Glasses, patching, or surgery41
Stroke-like episodes Low doses of aspirin42,43
Oral motor dysfunction Consider consultation with a gastroenterologist/nutritionist41
Osteopenia Bisphosphonate therapy in severe forms44,45
Pericardial effusion Corticosteroids and salicylic acid, pericardial drainage46
Scoliosis/kyphosis Surgical treatment in severe forms41
True hypothyroidism L-thyroxine supplementation41
3.5.3 Possible new treatment strategies. Treatment of PMM2-CDG

patients with Man1-P from external sources is ineffective due to its high
polarity, its inability to penetrate through the cellular membrane and its
instability in blood. Different chemically synthesized membrane-permeable
derivatives of Man-1-P have been produced in order to increase GDP-Man
pools. Some have been found to restore LLO biosynthesis and to increase
the level of GDP-Man in PMM2-CDG patients’ fibroblasts,4,47,48 but
further work has to be done in order to decrease their toxicity and enhance
their stability. Other types of CDG that rely on GDP-Man availability
might benefit as well from this approach.
Another study showed that metformin-stimulated D-mannose transport
(MSMT) increased mannose uptake by skin fibroblasts from PMM2-CDG
(CDG-Ia) patients increasing the LLO precursor pool. However, a ther-
apeutic administration of metformin is questionable because mannose by
itself is ineffective and the MSMT effect in other tissues is unknown.49,50
Recently, Dahl and colleagues51 discovered a series of potent inhibitors of
PMI with in vivo efficacy when combined with the mannose supplementa-
tion in disease models, as a strategy to redirect more Man-6-P towards the
depleted glycosylation pathway.
In a more recent study,50 the squalene synthase inhibitor Zaragozic acid
was reported to improve dolichol-linked oligosaccharide biosynthesis in
DPM1-CDG fibroblasts representing a therapeutic possibility for N-linked
glycosylation defects with residual enzymatic activity.
Furthermore, a report showed that functional and immunoreactive PMM
protein was obtained using antisense morpholino (AMO) in cultured
fibroblasts carrying three different splicing mutations in the PMM2 gene
proving to be a promising mutational-specific therapy option.32
The most common disease causing mutations in PMM2-CDG (CDGIa)
are missense changes which probably cause misfolding effects leading to
protein degradation. Recently, a remarkable study suggested that the loss-
of-function of several mutations in PMM2-CDG is based on their increased
susceptibility to degradation or aggregation.52 Therefore, pharmacological
chaperones (efficient in diseases such as phenylketonuria) may hold promise
for a large group of PMM2-CDG patients.53
3.6 CDG model systems

Biological model systems can contribute to a better understanding of the
pathophysiology of CDG and provide information needed to devise new
Carbohydr. Chem., 2012, 38, 124–155 | 141

Table 4 Summary of Biological model systems for CDGs.
Biological model
systems Advantages Disadvantages
Saccharomyces Rapid molecular diagnostic tool, Not useful for complex glycans
cerevisiae inexpensive, abundant growth typical of mammalian cells;
and good genetic analysis some proteins not present in
yeast (i.e. human LEC 35).
Golgi glycosylation pathways
differ between human and
yeast; unable to predict phy-
siological effects in patients
Mammalian cell lines Relatively easily manipulated and Do not reveal the physiological
transformed functions of glycans in multi-
cellular organisms
C. elegans Well characterized; small (1–1.5 Biochemistry and pathology are
mm); easy to cultivate; short difficult; cells post-reproduc-
life cycle (3 days); genome tive; lower invertebrate; many
completely sequenced; self- physiological functions not
fertilizing (hermaphrodite) conserved (immune system)64
Drosophila Short generation time (2 weeks); Life cycle longer than S. cerevi-
melanogaster cheap and easy to manipulate; siae and C. elegans; not useful
results extrapolable to humans for biological processes evolved
(many homologous disease in vertebrates
genes); established methods and
tools; suitable to model diseases
at the level of tissues or cell
communication65
Zebrafish Good to observe (small and Knock-out methods not available
(Danio rerio) transparent); relatively rapid yet (alternative use of Mor-
life cycle, easy breeding, grow- pholinos); many genetic tools
ing list of molecular tools66 still lacking
Mouse Short generation time (8–10 Excessive complexity; problems
weeks); easy to handle; low cost with embryonic lethality or
of maintenance; allows absence of phenotype, which
embryonic, physiological and may depend on the strain
therapeutical evaluation; high
identity between mice and
human orthologues
treatments that will have an impact on quality of life and life expectancy of
these patients. In addition, these models allow researchers to test ther-
apeutic strategies previously mentioned. Mammalian cell lines, Saccharo-
myces cerevisiae, C. elegans, zebrafish and mice have been studied in this
context (Table 4).
Saccharomyces cerevisiae has been an important model to elucidate the
molecular cause of several CDG types because there is a significant con-
servation of the N-linked glycosylation assembly between yeast and mam-
malian cells.54,55 Because Golgi glycosylation differs considerably between
yeast and animals, this model has been of little help to study Golgi-related
diseases.
Different mammalian cell lines with mutations affecting the biosynthesis
of N-linked oligosaccharides such as Chinese hamster ovary (CHO) cells,56
mouse BW5147 lymphoma cells and baby hamster kidney (BHK) are
142 | Carbohydr. Chem., 2012, 38, 124–155

currently available to test putative human mutations for which no yeast
ortholog exists.54,57
In an attempt to better understand CDG pathogenesis, Struwe and col-
leagues58,59 studied Caenorhabditis elegans to model N-linked glycosylation
defects. Several candidate genes have been identified as possibly responsible
for disease severity in humans.
Several other animal models have been important for the study of CDG.
For example, a study of zebrafish as a model for human SLC35C1-CDG
(CDGIIc) where it was shown for the first time that defects in protein
fucosylation lead to altered neuronal differentiation, maintenance, axon
branching and synapse formation.60 The gene responsible for this pathology
was found to encode a GDP-fucose transporter. It was found that the
Drosophila genome encodes at least another GDP-fucose transporter that is
involved in the O-fucosylation of Notch signaling.61
Mouse models provide an excellent tool to better understand the effects of
underglycosylation in an whole organism.8 In fact, different studies showed
that the complete loss of enzymatic activity by ablation of PMM2 or PMI
genes is lethal,54,57,62 highlighting the relevance of glycosylation for the
proper function of the organism from the beginning of the embryonic
development. Nearly all CDG patients appear to have variable degrees of
residual enzymatic activity. Therefore, the generation of hypomorphic
mouse models, is more likely to have a positive result by the introduction of
specific mutations in the gene of interest rather than by its total ablation.
Recently, a remarkable study used frequent mutations causing severe dis-
ease in PMM2-CDG (CDGIa) patients. It yielded embryos with a severe
phenotype that was reversible by feeding mannose to pregnant dams.63
Noteworthy, this model mimics the most common human CDG, and thus
holds great promise for the development of therapeutic strategies.
3.7 Socio-familial aspects for CDG patients

Non-profit patient organizations are growing rapidly as well as the number
of patients and the different types of rare diseases. Their mission mainly
comprises: support, education, research, awareness and advocacy. Their
participation in the decision-making process is being accompanied by a
greater awareness and different converging trends that are helping to
intensify their influence.
Problems faced by health professionals and CDG patients include:
(1) Limited knowledge of experts and support services; (2) Restricted access
to specialized health services or approved therapies; (3) Major lack of
information targeted to families about the ongoing basic and clinical
research; (4) Difficulties of providing advice, diagnosis and health inter-
ventions for these rare diseases; (5) Lack of (national and international)
financial resources; and (6) Need for better cooperation between research-
ers, medical doctors and patients.
CDG patient organizations aim to increase knowledge, awareness and
dissemination of CDG and foster cooperation among basic scientists,
clinicians and family advocacy groups.
Actions to reach this goal include coordination and organization of scientific
and medical communication activities, establishment of collaboration with
Carbohydr. Chem., 2012, 38, 124–155 | 143

key stakeholders such as Guı́a Metabólica (a powerful and interactive tool
targeted to families affected by inborn errors of metabolism); Fostering and
encouragement of research, inform and disseminate information amongst
families and experts, empowerment of the CDG community by enhancement
of understanding, training, awareness and advocacy for better quality of
life for patients and their families (at the national and international level).
Interestingly, several initiatives of CDG awareness and dissemination
have been established in partnership between CDG organizations,
researchers and clinicians such as the : (1) Practical guide targeted to CDG
families; (2) Fairy-tale: ‘‘Glycoland and the coloured antennas’’ (3) CDG
awareness and dissemination tool (4) CDG Online community (5) Inter-
national CDG logo (6) CDG community store, and others.
Our working model can be used by other Rare Metabolic Disorders.
4 Concluding remarks and future perspectives: CDG as a challenge for
glycobiomedicine in the next decade!
About 30 years ago the first patients with a CDG were described by Jaak
Jaeken. Following his ground-breaking discovery, a large number of pub-
lications and enormous advances have been made in our understanding of
CDG.
Several recent reviews cover the discovery of new types of CDG, the
clinical spectrum, animal models and advances towards a therapy. Amongst
the most relevant are certainly those that shed light on the importance of
improving and establishing effective tools for (early) diagnosis aiming for a
better quality of life outcomes of children and adults with CDG.
Here we focused on the number of significant tools that are currently
available in a growing number of expert clinical centres to screen and
validate CDG. The blood test to study the glycosylation status of trans-
ferrin continues to be the most frequent indicator of CDG. However,
technologies such as carbohydrate microarrays and glycan sequencing
profile will probably become more commonly used to dissect the role/
function of glycans, glycosylation pathway, and the properties of glyco-
proteins. Overall, glycans role in biology and in particular in this group of
pathologies will be deciphered.
Recent developments in diagnostic sequencing tools to elucidate the
molecular basis of genetic diseases, the exome sequencing, will contribute to
a faster diagnosis after the CDG clinical suspicion. However, to speed-up
diagnosis international collaborations involving multidisciplinary working
teams are necessary. They must include cell biologists, clinicians, molecular
biologists, biochemists, biostatisticians (rather than informatics) and others.
Initiatives such as the exome variant server (http://evs.gs.washington.edu/
EVS/), accurate phenotyping and functional validation of results using
different approaches such as new animal models, cellular phenotyping
(using human cells) and biochemical assays are an added value to a faster
and early diagnosis.
Recently, sequencing of individual genomes (whole exome sequencing)
became affordable and accessible for research and potentially for clinical
applications. Every experimental approach offers advantages and draw-
backs and special caution must be used in the attempt to examine data
144 | Carbohydr. Chem., 2012, 38, 124–155

obtained from this technique: ‘‘accidental’’ findings for other possible dis-
ease conditions will be obtained. It is important to remember that only few
CDGs are completely or partially treatable, as illustrated by the example of
PMI-CDG (CDG Ib) in which mannose supplementation is life saving.
Other therapeutic possibilities are on the horizon.
Patients suffer from the functional consequences of their mutations. Thus,
it is extremely important to understand the mechanisms that underlie the
genetic disease, and to devise better specific and symptomatic treatments.
The lack of knowledge on CDG often puts the life of patients at risk and
it is likely that many patients with a CDG have a wrong diagnosis leading to
multiple medical consultations and prescription of inappropriate or even
harmful drugs and treatments. Many patients are not receiving the appro-
priate treatment from physiotherapists, nutritionists, psychologists, etc. . .
CDG patients will benefit from the data pooled in the International CDG
patient registry as it allows a sufficient sample size for epidemiological
research, clinical surveillance, future treatment evaluation (efficacy) and
monitoring (safety), genotype-phenotype correlation and finally patient
quality of life outcomes.
Finally, international collaborations between affected families, clinicians
and researchers will continue to foster and enhance research in this field.
There is no doubt that Congenital Disorders of Glycosylation are an
exciting and expanding field that will attract the attention of many ‘‘gly-
cobiomedical doctors’’ or ‘‘carbohydrate-lovers’’ leading to promising dis-
coveries in the next decade.
5 Highlights
Congenital (genetic) disorders of glycosylation are a booming group of
diseases in which developments in genetics and glycan disciplines will lead to
exciting discoveries in the near future.
Development of diagnostic tools: in particular new CDG screening and
validation methods are needed.
Conventional and symptomatic treatments: their development should
receive special attention from clinicians.
Congenital Disorders of Glycosylation cut across many medical spe-
cialties: thus reference centers are urgently needed.
Development of animal models are important in understanding the
pathogenesis of these diseases. Recently, a mouse model for the most
common form of CDG was developed. Other animal models which still are
under-utilized in CDG will probably gain popularity, such as, zebrafish and
Drosophila.
Awareness and dissemination: Special attention should be given to
initiatives developed by CDG patients’ groups in partnership with clinicians
and researchers.
Patients’ registry: Optimization of the resources currently available
and consensus amongst researchers in order to continue pooling and
sharing data through the International CDG registry is highly recom-
mended. In addition, patient self-registration for the collection of data
might be helpful.
Carbohydr. Chem., 2012, 38, 124–155 | 145

To foster and encourage research more initiatives like ‘‘Horizon
2020’’, IRDiRC (International Consortium on Rare Disease Research),
E-Rare Call for Proposals 2012 or Sanco-Health funding are urgently need.
6 Further information
CDG patients and their interest groups are highly dynamic worldwide.
Their contact information is the following:
Portugal: http://sindromecdg.orgfree.com/
Spain: http://webs.ono.com/aescdg/
France: http://www.lesptitscdg.org/
USA: http://www.cdgfamilynetwork.org/
Canada: http://www.thefog.ca/
http://thelittlefightersfoundation.com/
Germany: https://www.cdg-syndrom.de/
Denmark: http://www.cdgforeningen.dk/
Sweden: http://www.cdgs.se/
If CDG is suspected, a serum transferring IZEF test should be performed.
In Europe and United states several referral centers offer diagnosis of CDG.
Abbreviations
AMO Antisense morpholino
Asn Asparagine
Arf1 ADP ribosylation factor 1
BFA Brefeldin A
BHK Baby hamster kidney
CDG Congenital Disorders of Glycosylation
CDT Carbohydrate-deficient transferrin
CTP Cytidine triphosphate
COG Conserved Oligomeric Golgi
CHO Chinese hamster ovary
CZE Capillary zone electrophoresis
Dol-P Dolichol phosphate
ER Endoplasmic reticulum
ES-MS Electrospray mass spectrometry
Fuc Fucose
Gal Galactose
GDP-Man Guanosinediphosphate-mannose
GlcNAc N-acetylglucosamine
Hk Hexokinase
HPLC High pressure liquid chromatography
Tf-IEF Isoelectrofocusing of serum transferrin
LLO Lipid-Linked Oligosaccharide
Man-6-P Mannose-6-phosphate
Man-1-P Mannose-1-phosphate
MSMT Metformin-stimulated D-mannose transport
OST Oligosaccharyltransferase
PMM2 Phosphomannomutase 2
146 | Carbohydr. Chem., 2012, 38, 124–155

PMI Phosphomannose isomerase
Ser Serine
NANA N-acetyl-neuraminic acid
Thr Threonine
UDP-GlcNAc Uridine-diphospho-N-acetylglucosamine
Acknowledgements
This chapter is dedicated to Dr. Jaak Jaeken (Leuven, Belgium) and to
Liliana Ferreira (Lisbon, Portugal). We thank Dr. Jaeken for his ongoing
scientific, medical and humane input which has contributed tremendously
to expanding CDG knowledge and awareness. His commitment to patient
organizations is exceptional and we think he is a successful model to follow.
We would like to thank Liliana Ferreira who is an inspiration and a driving
force for CDG patient advocacy.
We are also grateful to Dr. Gert Matthjis (Leuven, Belgium) for his
dedication to CDG research at the European level: Euroglycan and Euro-
glycanet greatly improved CDG (early) diagnosis and established successful
international collaborations amongst clinical and basic research groups. We
would like to express our gratitude also to all families and CDG experts
whose encouragement and support for CDG patient activities enabled the
development of many projects in a productive partnership. The final goal is
to expand CDG research and to improve patients’ disease outcome.
References
1 R. Apweiler, H. Hermjakob and N. Sharon, Biochim Biophys Acta, 1999, 1473,
4–8.
2 H. Schachter and H. H. Freeze, Biochim Biophys Acta, 2009, 1792, 925–930.
3 J. S. Rush, N. Gao, M. A. Lehrman and C. J. Waechter, J Biol Chem, 2008,
283, 4087–4093.
4 R. Hardre, A. Khaled, A. Willemetz, T. Dupre, S. Moore, C. Gravier-Pelletier
and Y. Le Merrer, Bioorg Med Chem Lett, 2007, 17, 152–155.
5 T. Marquardt and J. Denecke, Eur J Pediatr, 2003, 162, 359–379.
6 D. J. Kelleher and R. Gilmore, Glycobiology, 2006, 16, 47R–62R.
7 E. Mohorko, R. Glockshuber and M. Aebi, J Inherit Metab Dis, 34, 869–878.
8 C. Thiel and C. Korner, J Inherit Metab Dis, 34, 879–889.
9 S. Wopereis, D. J. Lefeber, E. Morava and R. A. Wevers, Clin Chem, 2006, 52,
574–600.
10 W. Vleugels, E. Schollen, F. Foulquier and G. Matthijs, Biochem Biophys Res
Commun, 2009, 390, 769–774.
11 M. Mohamed, D. Kouwenberg, T. Gardeitchik, U. Kornak, R. A. Wevers and
E. Morava, J Inherit Metab Dis, 2011, 34, 907–916.
12 J. Jaeken, J Inherit Metab Dis, 2011, 34, 853–858.
13 F. Foulquier, E. Vasile, E. Schollen, N. Callewaert, T. Raemaekers, D.
Quelhas, J. Jaeken, P. Mills, B. Winchester, M. Krieger, W. Annaert and G.
Matthijs, Proc Natl Acad Sci USA, 2006, 103, 3764–3769.
14 F. Foulquier, D. Ungar, E. Reynders, R. Zeevaert, P. Mills, M. T. Garcia-
Silva, P. Briones, B. Winchester, W. Morelle, M. Krieger, W. Annaert and G.
Matthijs, Hum Mol Genet, 2007, 16, 717–730.
15 E. Reynders, F. Foulquier, E. Leao Teles, D. Quelhas, W. Morelle, C.
Rabouille, W. Annaert and G. Matthijs, Hum Mol Genet, 2009, 18, 3244–3256.
Carbohydr. Chem., 2012, 38, 124–155 | 147

16 R. Zeevaert, F. Foulquier, D. Cheillan, I. Cloix, N. Guffon, L. Sturiale, D.
Garozzo, G. Matthijs and J. Jaeken, Eur J Med Genet, 2009, 52, 303–305.
17 C. Kranz, B. G. Ng, L. Sun, V. Sharma, E. A. Eklund, Y. Miura, D. Ungar, V.
Lupashin, R. D. Winkel, J. F. Cipollo, C. E. Costello, E. Loh, W. Hong and
H. H. Freeze, Hum Mol Genet, 2007, 16, 731–741.
18 P. Paesold-Burda, C. Maag, H. Troxler, F. Foulquier, P. Kleinert, S. Schnabel,
M. Baumgartner and T. Hennet, Hum Mol Genet, 2009, 18, 4350–4356.
19 X. Wu, R. A. Steet, O. Bohorov, J. Bakker, J. Newell, M. Krieger, L. Spaapen,
S. Kornfeld and H. H. Freeze, Nat Med, 2004, 10, 518–523.
20 M. Aebi, A. Helenius, B. Schenk, R. Barone, A. Fiumara, E. G. Berger, T.
Hennet, T. Imbach, A. Stutz, C. Bjursell, A. Uller, J. G. Wahlstrom, P.
Briones, E. Cardo, P. Clayton, B. Winchester, V. Cormier-Dalre, P. de Lonlay,
M. Cuer, T. Dupre, N. Seta, T. de Koning, L. Dorland, F. de Loos and L.
Kupers, et al., Glycoconj J, 1999, 16, 669–671.
21 J. Jaeken, T. Hennet, H. H. Freeze and G. Matthijs, J Inherit Metab Dis, 2008,
31, 669–672.
22 J. Jaeken, T. Hennet, G. Matthijs and H. H. Freeze, Biochim Biophys Acta,
2009, 1792, 825–826.
23 J. Jaeken, J Inherit Metab Dis, 2003, 26, 99–118.
24 P. Burda, L. Borsig, J. de Rijk-van Andel, R. Wevers, J. Jaeken, H. Carchon,
E. G. Berger and M. Aebi, J Clin Invest, 1998, 102, 647–652.
25 S. Wopereis, E. Morava, S. Grunewald, M. Adamowicz, K. M. Huijben, D. J.
Lefeber and R. A. Wevers, Glycobiology, 2005, 15, 1312–1319.
26 G. Matthijs, E. Schollen, E. Van Schaftingen, J. J. Cassiman and J. Jaeken,
Am J Hum Genet, 1998, 62, 542–550.
27 C. Le Bizec, S. Vuillaumier-Barrot, A. Barnier, T. Dupre, G. Durand and
N. Seta, Hum Mutat, 2005, 25, 504–505.
28 S. Kjaergaard, F. Skovby and M. Schwartz, Eur J Hum Genet, 1998, 6, 331–336.
29 P. Briones, M. A. Vilaseca, E. Schollen, I. Ferrer, M. Maties, C. Busquets, R.
Artuch, L. Gort, M. Marco, E. van Schaftingen, G. Matthijs, J. Jaeken and A.
Chabas, J Inherit Metab Dis, 2002, 25, 635–646.
30 M. Pirard, G. Matthijs, L. Heykants, E. Schollen, S. Grunewald, J. Jaeken and
E. van Schaftingen, FEBS Lett, 1999, 452, 319–322.
31 S. Kjaergaard, F. Skovby and M. Schwartz, Eur J Hum Genet, 1999, 7,
884–888.
32 A. I. Vega, C. Perez-Cerda, L. R. Desviat, G. Matthijs, M. Ugarte and B.
Perez, Hum Mutat, 2009, 30, 795–803.
33 C. J. Hendriksz, P. McClean, M. J. Henderson, D. G. Keir, V. C.
Worthington, F. Imtiaz, E. Schollen, G. Matthijs and B. G. Winchester, Arch
Dis Child, 2001, 85, 339–340.
34 H. K. Harms, K. P. Zimmer, K. Kurnik, R. M. Bertele-Harms, S. Weidinger
and K. Reiter, Acta Paediatr, 2002, 91, 1065–1072.
35 R. Niehues and M. Hasilik, MMW Fortschr Med, 2000, 142, 171–172.
36 V. Westphal, S. Kjaergaard, J. A. Davis, S. M. Peterson, F. Skovby and
H. H. Freeze, Mol Genet Metab, 2001, 73, 77–85.
37 E. Martin Hernandez, A. I. Vega Pajares, B. Perez Gonzalez, M. J. Ecay
Crespo, F. Leal Perez, J. Manzanares Lopez-Manzanares, M. Ugarte Perez
and C. Perez-Cerda Silvestre, An Pediatr (Barc), 2008, 69, 358–365.
38 R. Y. Tamminga, D. J. Lefeber, W. A. Kamps and F. J. van Spronsen, Pediatr
Hematol Oncol, 2008, 25, 762–768.
39 P. de Lonlay and N. Seta, Biochim Biophys Acta, 2009, 1792, 841–843.
40 A. Almeida, M. Layton and A. Karadimitris, Biochim Biophys Acta, 2009,
1792, 874–880.
148 | Carbohydr. Chem., 2012, 38, 124–155

41 S. E. Sparks and D. M. Krasnewich, 1993.
42 J. B. Arnoux, N. Boddaert, V. Valayannopoulos, S. Romano, N. Bahi-
Buisson, I. Desguerre, Y. de Keyzer, A. Munnich, F. Brunelle, N. Seta, M. D.
Dautzenberg and P. de Lonlay, Mol Genet Metab, 2008, 93, 444–449.
43 N. Ishikawa, G. Tajima, H. Ono and M. Kobayashi, Brain Dev, 2009, 31,
240–243.
44 D. Coman, M. Irving, P. Kannu, J. Jaeken and R. Savarirayan, Clin Genet,
2008, 73, 507–515.
45 D. Coman, D. Bostock, M. Hunter, P. Kannu, M. Irving, V. Mayne, M. Fietz,
J. Jaeken and R. Savarirayan, Am J Med Genet A, 2008, 146, 389–392.
46 B. J. Feldman and D. Rosenthal, Pediatr Cardiol, 2002, 23, 469–471.
47 S. Rutschow, J. Thiem, C. Kranz and T. Marquardt, Bioorg Med Chem, 2002,
10, 4043–4049.
48 E. A. Eklund, N. Merbouh, M. Ichikawa, A. Nishikawa, J. M. Clima, J. A.
Dorman, T. Norberg and H. H. Freeze, Glycobiology, 2005, 15, 1084–1093.
49 J. Shang and M. A. Lehrman, J Biol Chem, 2004, 279, 9703–9712.
50 M. A. Haeuptle, M. Welti, H. Troxler, A. J. Hulsmeier, T. Imbach and T.
Hennet, J Biol Chem, 286, 6085–6091.
51 R. Dahl, Y. Bravo, V. Sharma, M. Ichikawa, R. P. Dhanya, M. Hedrick, B.
Brown, J. Rascon, M. Vicchiarelli, A. Mangravita-Novo, L. Yang, D. Stonich,
Y. Su, L. H. Smith, E. Sergienko, H. H. Freeze and N. D. Cosford, J Med
Chem, 54, 3661–3668.
52 A. I. Vega, C. Perez-Cerda, D. Abia, A. Gamez, P. Briones, R. Artuch, L. R.
Desviat, M. Ugarte and B. Perez, J Inherit Metab Dis, 34, 929–939.
53 H. H. Freeze, Biochim Biophys Acta, 2009, 1792, 835–840.
54 T. Marquardt and H. Freeze, Biol Chem, 2001, 382, 161–177.
55 M. Aebi and T. Hennet, Trends Cell Biol, 2001, 11, 136–141.
56 M. B. Jones, N. Tomiya, M. J. Betenbaugh and S. S. Krag, Biochem Biophys
Res Commun, 395, 36–41.
57 H. H. Freeze, Biochim Biophys Acta, 2002, 1573, 388–393.
58 W. B. Struwe, B. L. Hughes, D. W. Osborn, E. D. Boudreau, K. M. Shaw and
C. E. Warren, Glycobiology, 2009, 19, 1554–1562.
59 W. B. Struwe and C. E. Warren, Methods Enzymol, 480, 477–493.
60 Y. Song, J. R. Willer, P. C. Scherer, J. A. Panzer, A. Kugath, E. Skordalakes,
R. G. Gregg, G. B. Willer and R. J. Balice-Gordon, PLoS One, 2010, 5,
e13743.
61 H. O. Ishikawa, S. Higashi, T. Ayukawa, T. Sasamura, M. Kitagawa, K.
Harigaya, K. Aoki, N. Ishida, Y. Sanai and K. Matsuno, Proc Natl Acad Sci
USA, 2005, 102, 18532–18537.
62 C. Thiel, T. Lubke, G. Matthijs, K. von Figura and C. Korner, Mol Cell Biol,
2006, 26, 5615–5620.
63 A. Schneider, C. Thiel, J. Rindermann, C. Derossi, D. Popovici, G. F.
Hoffmann, H. J. Grone and C. Korner, Nat Med, 2011, 18, 71–73.
64 T. E. Johnson, Exp Gerontol, 2003, 38, 1329–1332.
65 E. Bier, Nat Rev Genet, 2005, 6, 9–23.
66 B. Novoa and A. Figueras, Adv Exp Med Biol, 946, 253–275.
67 P. de Lonlay, V. Valayannopoulos, T. Dupre, S. Vuillaumier-Barrot and N.
Seta, Arch Pediatr, 2008, 15, 602–605.
68 J. Jaeken and G. Matthijs, Annu Rev Genomics Hum Genet, 2007, 8, 261–278.
69 J. Jaeken, Ann N Y Acad Sci, 2010, 1214, 190–198.
70 E. Morava and D. Lefeber, J Inherit Metab Dis, 2011, 34, 847–848.
71 D. J. Lefeber, E. Morava and J. Jaeken, J Inherit Metab Dis, 2011, 34, 849–852.
72 M. A. Haeuptle and T. Hennet, Hum Mutat, 2009, 30, 1628–1641.
Carbohydr. Chem., 2012, 38, 124–155 | 149

73 S. E. Sparks and D. M. Krasnewich, 1993.
74 E. Van Schaftingen and J. Jaeken, FEBS Lett, 1995, 377, 318–320.
75 B. Kristiansson, H. Stibler, B. Hagberg and J. Wahlstrom, Lakartidningen,
1998, 95, 5742–5748.
76 H. Carchon, E. Van Schaftingen, G. Matthijs and J. Jaeken, Biochim Biophys
Acta, 1999, 1455, 155–165.
77 C. H. van Ommen, M. Peters, P. G. Barth, P. Vreken, R. J. Wanders and J.
Jaeken, J Pediatr, 2000, 136, 400–403.
78 T. Dupre, M. Cuer, S. Barrot, A. Barnier, V. Cormier-Daire, A. Munnich, G.
Durand and N. Seta, Clin Chem, 2001, 47, 132–134.
79 S. Kjaergaard, M. Schwartz and F. Skovby, Arch Dis Child, 2001, 85,
236–239.
80 S. Kjaergaard, J. Muller and F. Skovby, Arch Dis Child, 2002, 87, 324–327.
81 I. Mader, M. Dobler-Neumann, W. Kuker, H. Stibler and I. Krageloh-Mann,
Childs Nerv Syst, 2002, 18, 77–80.
82 G. Sala, T. Dupre, N. Seta, P. Codogno and R. Ghidoni, Pediatr Res, 2002, 52,
645–651.
83 T. Honzik, E. Malonova, H. Hansikova, R. Rosipal, H. Poupetova, P.
Martasek and J. Zeman, Cas Lek Cesk, 2003, 142, 276–279.
84 L. M. Neumann, A. von Moers, J. Kunze, O. Blankenstein and T. Marquardt,
Eur J Pediatr, 2003, 162, 710–713.
85 K. L. Chu, Y. H. Chien, C. E. Tsai, H. H. Freeze, E. Eklund and W. L. Hwu,
J Formos Med Assoc, 2004, 103, 721–723.
86 E. Aronica, A. A. van Kempen, M. van der Heide, B. T. Poll-The, H. J. van
Slooten, D. Troost and J. M. Rozemuller-Kwakkel, Acta Neuropathol, 2005,
109, 433–442.
87 I. Giurgea, A. Michel, M. Le Merrer, N. Seta and P. de Lonlay, Pediatr
Neurol, 2005, 32, 121–123.
88 D. Coman, J. McGill, R. MacDonald, D. Morris, S. Klingberg, J. Jaeken and
D. Appleton, J Clin Neurosci, 2007, 14, 668–672.
89 D. Krasnewich, K. O’Brien and S. Sparks, Am J Med Genet C Semin Med
Genet, 2007, 145C, 302–306.
90 D. Wurm, A. Hansgen, Y. J. Kim, A. Lindinger, A. Baghai and L. Gortner,
Eur J Pediatr, 2007, 166, 377–378.
91 D. Wurm, G. Loffler, A. Lindinger and L. Gortner, J Perinatol, 2007, 27,
802–804.
92 R. Barone, L. Sturiale, V. Sofia, A. Ignoto, A. Fiumara, G. Sorge, D. Garozzo
and M. Zappia, Am J Med Genet A, 2008, 146A, 2103–2108.
93 J. M. Brum, I. M. Rizzo, W. D. Mello and C. E. Speck-Martins, Arq Neu-
ropsiquiatr, 2008, 66, 545–548.
94 S. Grunewald, Biochim Biophys Acta, 2009, 1792, 827–834.
95 E. Jamroz, D. Adamek, J. Paprocka, M. Adamowicz, E. Marszal and R. A.
Wevers, J Child Neurol, 2009, 24, 13–18.
96 B. Perez-Duenas, A. Garcia-Cazorla, M. Pineda, P. Poo, J. Campistol, V.
Cusi, E. Schollen, G. Matthijs, S. Grunewald, P. Briones, C. Perez-Cerda, R.
Artuch and M. A. Vilaseca, Eur J Paediatr Neurol, 2009, 13, 444–451.
97 M. K. Thong, M. Fietz, C. Nicholls, M. H. Lee and O. Asma, J Inherit Metab
Dis, 2009.
98 J. Jaeken, G. Matthijs, J. M. Saudubray, C. Dionisi-Vici, E. Bertini, P. de
Lonlay, H. Henri, H. Carchon, E. Schollen and E. Van Schaftingen, Am J
Hum Genet, 1998, 62, 1535–1539.
99 E. Schollen, L. Dorland, T. J. de Koning, O. P. Van Diggelen, J. G. Huijmans,
T. Marquardt, D. Babovic-Vuksanovic, M. Patterson, F. Imtiaz, B.
150 | Carbohydr. Chem., 2012, 38, 124–155

Winchester, M. Adamowicz, E. Pronicka, H. Freeze and G. Matthijs, Hum
Mutat, 2000, 16, 247–252.
100 D. F. Kelly, A. Boneh, S. Pitsch, H. Gold, M. Fietz, P. Nelson and M. R.
Oliver, J Paediatr Child Health, 2001, 37, 510–512.
101 D. Penel-Capelle, D. Dobbelaere, J. Jaeken, A. Klein, M. Cartigny and J.
Weill, J Inherit Metab Dis, 2003, 26, 83–85.
102 M. Al-Owain, S. Mohamed, N. Kaya, A. Zagal, G. Matthijs and J. Jaeken,
Orphanet J Rare Dis, 2010, 5, 7.
103 B. S. Miller, H. H. Freeze, G. F. Hoffmann and K. Sarafoglou, Mol Genet
Metab, 2011, 103, 101–103.
104 T. Imbach, P. Burda, P. Kuhnert, R. A. Wevers, M. Aebi, E. G. Berger and T.
Hennet, Proc Natl Acad Sci USA, 1999, 96, 6982–6987.
105 S. Grunewald, T. Imbach, K. Huijben, M. E. Rubio-Gozalbo, A. Verrips, J. B.
de Klerk, H. Stroink, J. F. de Rijk-van Andel, J. L. Van Hove, U. Wendel, G.
Matthijs, T. Hennet, J. Jaeken and R. A. Wevers, Ann Neurol, 2000, 47, 776–781.
106 F. Hanefeld, C. Korner, U. Holzbach-Eberle and K. von Figura, Neurope-
diatrics, 2000, 31, 60–62.
107 T. Imbach, S. Grunewald, B. Schenk, P. Burda, E. Schollen, R. A. Wevers, J.
Jaeken, J. B. de Klerk, E. G. Berger, G. Matthijs, M. Aebi and T. Hennet,
Hum Genet, 2000, 106, 538–545.
108 V. Westphal, C. Schottstadt, T. Marquardt and H. H. Freeze, Mol Genet
Metab, 2000, 70, 219–223.
109 S. Vuillaumier-Barrot, C. Le Bizec, G. Durand and N. Seta, J Hum Genet,
2001, 46, 547–548.
110 J. W. Newell, N. S. Seo, G. M. Enns, M. McCraken, J. F. Mantovani and H.
H. Freeze, Mol Genet Metab, 2003, 79, 221–228.
111 V. Westphal, M. Xiao, P. Y. Kwok and H. H. Freeze, Hum Mutat, 2003, 22,
420–421.
112 L. Sun, E. A. Eklund, J. L. Van Hove, H. H. Freeze and J. A. Thomas, Am J
Med Genet A, 2005, 137, 22–26.
113 J. Denecke, C. Kranz, J. von Kleist-Retzow, K. Bosse, P. Herkenrath, O.
Debus, E. Harms and T. Marquardt, Pediatr Res, 2005, 58, 248–253.
114 E. Schollen, S. Grunewald, L. Keldermans, B. Albrecht, C. Korner and G.
Matthijs, Eur J Med Genet, 2005, 48, 153–158.
115 C. Kranz, L. Sun, E. A. Eklund, D. Krasnewich, J. R. Casey and H. H. Freeze,
Am J Med Genet A, 2007, 143A, 1414–1420.
116 A. Rimella-Le-Huu, H. Henry, I. Kern, S. Hanquinet, E. Roulet-Perez, C. J.
Newman, A. Superti-Furga, L. Bonafe and D. Ballhausen, J Inherit Metab
Dis, 2008.
117 I. Chantret, T. Dupre, C. Delenda, S. Bucher, J. Dancourt, A. Barnier,
A. Charollais, D. Heron, B. Bader-Meunier, O. Danos, N. Seta, G. Durand,
R. Oriol, P. Codogno and S. E. Moore, J Biol Chem, 2002, 277, 25815–
25822.
118 C. E. Grubenmann, C. G. Frank, S. Kjaergaard, E. G. Berger, M. Aebi and T.
Hennet, Hum Mol Genet, 2002, 11, 2331–2339.
119 M. Di Rocco, T. Hennet, C. E. Grubenmann, S. Pagliardini, A. E. Allegri, C.
G. Frank, M. Aebi, S. Vignola and J. Jaeken, J Inherit Metab Dis, 2005, 28,
1162–1164.
120 E. A. Eklund, J. W. Newell, L. Sun, N. S. Seo, G. Alper, J. Willert and H. H.
Freeze, Mol Genet Metab, 2005, 84, 25–31.
121 C. Kranz, A. A. Basinger, M. Gucsavas-Calikoglu, L. Sun, C. M. Powell, F.
W. Henderson, A. S. Aylsworth and H. H. Freeze, Am J Med Genet A, 2007,
143A, 1371–1378.
Carbohydr. Chem., 2012, 38, 124–155 | 151

122 D. Peric, C. Durrant-Arico, C. Delenda, T. Dupre, P. De Lonlay, H. O. de
Baulny, C. Pelatan, B. Bader-Meunier, O. Danos, I. Chantret and S. E.
Moore, PLoS One, 2010, 5, e11675.
123 I. Chantret, J. Dancourt, T. Dupre, C. Delenda, S. Bucher, S. Vuillaumier-
Barrot, H. Ogier de Baulny, C. Peletan, O. Danos, N. Seta, G. Durand, R.
Oriol, P. Codogno and S. E. Moore, J Biol Chem, 2003, 278, 9962–9971.
124 E. Schollen, C. G. Frank, L. Keldermans, R. Reyntjens, C. E. Grubenmann, P.
T. Clayton, B. G. Winchester, J. Smeitink, R. A. Wevers, M. Aebi, T. Hennet
and G. Matthijs, J Med Genet, 2004, 41, 550–556.
125 E. A. Eklund, L. Sun, V. Westphal, J. L. Northrop, H. H. Freeze and F.
Scaglia, J Pediatr, 2005, 147, 847–850.
126 M. Jackson, P. Clayton, S. Grunewald, G. Keir, K. Mills, P. Mills, B.
Winchester, V. Worthington and E. Young, J Inherit Metab Dis, 2005, 28,
1197–1198.
127 T. Stolting, H. Omran, A. Erlekotte, J. Denecke, J. Reunert and T.
Marquardt, Mol Genet Metab, 2009, 98, 305–309.
128 K. Vesela, T. Honzik, H. Hansikova, M. A. Haeuptle, J. Semberova, Z.
Stranak, T. Hennet and J. Zeman, J Inherit Metab Dis, 2009.
129 C. Thiel, M. Schwarz, J. Peng, M. Grzmil, M. Hasilik, T. Braulke, A.
Kohlschutter, K. von Figura, L. Lehle and C. Korner, J Biol Chem, 2003, 278,
22498–22505.
130 X. Wu, J. S. Rush, D. Karaoglu, D. Krasnewich, M. S. Lubinsky, C. J.
Waechter, R. Gilmore and H. H. Freeze, Hum Mutat, 2003, 22, 144–150.
131 T. Dupre, S. Vuillaumier-Barrot, I. Chantret, H. S. Yaye, C. Le Bizec, A.
Afenjar, C. Altuzarra, C. Barnerias, L. Burglen, P. de Lonlay, F. Feillet, S.
Napuri, N. Seta and S. E. Moore, J Med Genet, 2010, 47, 729–735.
132 C. E. Grubenmann, C. G. Frank, A. J. Hulsmeier, E. Schollen, G. Matthijs, E.
Mayatepek, E. G. Berger, M. Aebi and T. Hennet, Hum Mol Genet, 2004, 13,
535–542.
133 C. Kranz, J. Denecke, L. Lehle, K. Sohlbach, S. Jeske, F. Meinhardt, R. Rossi,
S. Gudowius and T. Marquardt, Am J Hum Genet, 2004, 74, 545–551.
134 M. Schwarz, C. Thiel, J. Lubbehusen, B. Dorland, T. de Koning, K. von
Figura, L. Lehle and C. Korner, Am J Hum Genet, 2004, 74, 472–481.
135 C. G. Frank, C. E. Grubenmann, W. Eyaid, E. G. Berger, M. Aebi and T.
Hennet, Am J Hum Genet, 2004, 75, 146–150.
136 M. Weinstein, E. Schollen, G. Matthijs, C. Neupert, T. Hennet, C. E.
Grubenmann, C. G. Frank, M. Aebi, J. T. Clarke, A. Griffiths, L. Seargeant
and N. Poplawski, Am J Med Genet A, 2005, 136, 194–197.
137 W. Vleugels, L. Keldermans, J. Jaeken, T. D. Butters, J. C. Michalski, G.
Matthijs and F. Foulquier, Glycobiology, 2009, 19, 910–917.
138 B. E. Baysal, J. E. Willett-Brozick, S. A. Bacanu, S. Detera-Wadleigh and V.
L. Nimgaonkar, Behav Brain Funct, 2006, 2, 25.
139 J. Jaeken, W. Vleugels, L. Regal, C. Corchia, N. Goemans, M. A. Haeuptle,
F. Foulquier, T. Hennet, G. Matthijs and C. Dionisi-Vici, J Inherit Metab Dis,
2009.
140 W. Vleugels, M. A. Haeuptle, B. G. Ng, J. C. Michalski, R. Battini, C.
Dionisi-Vici, M. D. Ludman, J. Jaeken, F. Foulquier, H. H. Freeze, G.
Matthijs and T. Hennet, Hum Mutat, 2009, 30, 1428–1434.
141 N. Rind, V. Schmeiser, C. Thiel, B. Absmanner, J. Lubbehusen, J. Hocks, N.
Apeshiotis, E. Wilichowski, L. Lehle and C. Korner, Hum Mol Genet, 2010,
19, 1413–1424.
142 B. B. de Vries, W. G. van’tHoff, R. A. Surtees and R. M. Winter, Clin
Dysmorphol, 2001, 10, 115–121.
152 | Carbohydr. Chem., 2012, 38, 124–155

143 C. Van Geet, J. Jaeken, K. Freson, T. Lenaerts, J. Arnout, J. Vermylen and M.
F. Hoylaerts, J Inherit Metab Dis, 2001, 24, 477–492.
144 P. de Cock and J. Jaeken, Biochim Biophys Acta, 2009, 1792, 844–846.
145 C. Volker, C. M. De Praeter, B. Hardt, W. Breuer, B. Kalz-Fuller, R. N. Van
Coster and E. Bause, Glycobiology, 2002, 12, 473–483.
146 M. Garshasbi, K. Kahrizi, M. Hosseini, L. Nouri Vahid, M. Falah, S.
Hemmati, H. Hu, A. Tzschach, H. H. Ropers, H. Najmabadi and A. W. Kuss,
Am J Med Genet A, 2011, 155A, 1976–1980.
147 M. A. Khan, M. A. Rafiq, A. Noor, N. Ali, G. Ali, J. B. Vincent and M.
Ansar, BMC Med Genet, 2011, 12, 56.
148 E. Mohorko, R. Glockshuber and M. Aebi, J Inherit Metab Dis, 2011, 34, 869–
878.
149 M. Garshasbi, V. Hadavi, H. Habibi, K. Kahrizi, R. Kariminejad, F. Behjati,
A. Tzschach, H. Najmabadi, H. H. Ropers and A. W. Kuss, Am J Hum Genet,
2008, 82, 1158–1164.
150 F. Molinari, F. Foulquier, P. S. Tarpey, W. Morelle, S. Boissel,
J. Teague, S. Edkins, P. A. Futreal, M. R. Stratton, G. Turner, G. Matthijs,
J. Gecz, A. Munnich and L. Colleaux, Am J Hum Genet, 2008, 82,
1150–1157.
151 M. K. Stancheva-Ivanova, W. Wuyts, E. van Hul, B. I. Radeva, R. V.
Vazharova, T. P. Sokolov, B. Y. Vladimirov, M. D. Apostolova and I. M.
Kremensky, J Inherit Metab Dis, 34, 917–921.
152 A. De Paepe and F. Malfait, Clin Genet, 2012.
153 S. Ichikawa, G. Baujat, A. Seyahi, A. G. Garoufali, E. A. Imel, L. R. Padgett,
A. M. Austin, A. H. Sorenson, Z. Pejin, V. Topouchian, P. Quartier, V.
Cormier-Daire, M. Dechaux, F. Malandrinou, P. N. Singhellakis, M.
Le Merrer and M. J. Econs, Am J Med Genet A, 2010, 152A, 896–903.
154 A. M. Shenoy, J. A. Markowitz, C. G. Bonnemann, K. Krishnamoorthy,
A. D. Bossler and B. S. Tseng, J Clin Neuromuscul Dis, 2010, 11,
124–126.
155 T. Y. Heinonen and M. Maki, Ann Med, 2009, 41, 2–10.
156 D. Hess, J. J. Keusch, S. A. Oberstein, R. C. Hennekam and J. Hofsteenge,
J Biol Chem, 2008, 283, 7354–7360.
157 S. Hiraoka, T. Furuichi, G. Nishimura, S. Shibata, M. Yanagishita, D. L.
Rimoin, A. Superti-Furga, P. G. Nikkels, M. Ogawa, K. Katsuyama, H.
Toyoda, A. Kinoshita-Toyoda, N. Ishida, K. Isono, Y. Sanai, D. H. Cohn, H.
Koseki and S. Ikegawa, Nat Med, 2007, 13, 1363–1367.
158 T. Imbach, B. Schenk, E. Schollen, P. Burda, A. Stutz, S. Grunewald, N. M.
Bailie, M. D. King, J. Jaeken, G. Matthijs, E. G. Berger, M. Aebi and T.
Hennet, J Clin Invest, 2000, 105, 233–239.
159 S. Kim, V. Westphal, G. Srikrishna, D. P. Mehta, S. Peterson, J. Filiano,
P. S. Karnes, M. C. Patterson and H. H. Freeze, J Clin Invest, 2000, 105,
191–198.
160 C. Kranz, J. Denecke, M. A. Lehrman, S. Ray, P. Kienz, G. Kreissel, D. Sagi,
J. Peter-Katalinic, H. H. Freeze, T. Schmid, S. Jackowski-Dohrmann, E.
Harms and T. Marquardt, J Clin Invest, 2001, 108, 1613–1619.
161 B. Schenk, T. Imbach, C. G. Frank, C. E. Grubenmann, G. V. Raymond, H.
Hurvitz, I. Korn-Lubetzki, S. Revel-Vik, A. Raas-Rotschild, A. S. Luder, J.
Jaeken, E. G. Berger, G. Matthijs, T. Hennet and M. Aebi, J Clin Invest, 2001,
108, 1687–1695.
162 M. T. Garcia-Silva, G. Matthijs, E. Schollen, J. C. Cabrera, J. Sanchez del
Pozo, M. Marti Herreros, R. Simon, M. Maties, E. Martin Hernandez, T.
Hennet and P. Briones, J Inherit Metab Dis, 2004, 27, 591–600.
Carbohydr. Chem., 2012, 38, 124–155 | 153

163 J. Dancourt, S. Vuillaumier-Barrot, H. O. de Baulny, I. Sfaello, A. Barnier, C.
le Bizec, T. Dupre, G. Durand, N. Seta and S. E. Moore, Pediatr Res, 2006, 59,
835–839.
164 D. J. Lefeber, J. Schonberger, E. Morava, M. Guillard, K. M. Huyben, K.
Verrijp, O. Grafakou, A. Evangeliou, F. W. Preijers, P. Manta, J. Yildiz, S.
Grunewald, M. Spilioti, C. van den Elzen, D. Klein, D. Hess, H. Ashida, J.
Hofsteenge, Y. Maeda, L. van den Heuvel, M. Lammens, L. Lehle and R. A.
Wevers, Am J Hum Genet, 2009, 85, 76–86.
165 A. Broccolini, T. Gidaro, R. Morosetti, C. Sancricca and M. Mirabella, Acta
Myol, 2011, 30, 91–95.
166 T. Lubke, T. Marquardt, K. von Figura and C. Korner, J Biol Chem, 1999,
274, 25986–25989.
167 T. Lubke, T. Marquardt, A. Etzioni, E. Hartmann, K. von Figura and C.
Korner, Nat Genet, 2001, 28, 73–76.
168 A. Etzioni, L. Sturla, A. Antonellis, E. D. Green, R. Gershoni-Baruch, P. M.
Berninsone, C. B. Hirschberg and M. Tonetti, Am J Med Genet, 2002, 110,
131–135.
169 N. Ishida and M. Kawakita, Pflugers Arch, 2004, 447, 768–775.
170 Y. Luo and R. S. Haltiwanger, J Biol Chem, 2005, 280, 11289–11294.
171 L. Sturla, F. Fruscione, K. Noda, E. Miyoshi, N. Taniguchi, P. Contini and
M. Tonetti, Glycobiology, 2005, 15, 924–934.
172 C. C. Hellbusch, M. Sperandio, D. Frommhold, S. Yakubenia, M. K. Wild, D.
Popovici, D. Vestweber, H. J. Grone, K. von Figura, T. Lubke and C. Korner,
J Biol Chem, 2007, 282, 10762–10772.
173 B. Hansske, C. Thiel, T. Lubke, M. Hasilik, S. Honing, V. Peters, P. H.
Heidemann, G. F. Hoffmann, E. G. Berger, K. von Figura and C. Korner, J
Clin Invest, 2002, 109, 725–733.
174 V. Peters, J. M. Penzien, G. Reiter, C. Korner, R. Hackler, B. Assmann, J.
Fang, J. R. Schaefer, G. F. Hoffmann and P. H. Heidemann, Neuropediatrics,
2002, 33, 27–32.
175 C. Kranz, C. Jungeblut, J. Denecke, A. Erlekotte, C. Sohlbach, V. Debus, H.
G. Kehl, E. Harms, A. Reith, S. Reichel, H. Grobe, G. Hammersen, U.
Schwarzer and T. Marquardt, Am J Hum Genet, 2007, 80, 433–440.
176 V. Cantagrel, D. J. Lefeber, B. G. Ng, Z. Guan, J. L. Silhavy, S. L. Bielas, L.
Lehle, H. Hombauer, M. Adamowicz, E. Swiezewska, A. P. De Brouwer, P.
Blumel, J. Sykut-Cegielska, S. Houliston, D. Swistun, B. R. Ali, W. B.
Dobyns, D. Babovic-Vuksanovic, H. van Bokhoven, R. A. Wevers, C. R.
Raetz, H. H. Freeze, E. Morava, L. Al-Gazali and J. G. Gleeson, Cell, 2010,
142, 203–217.
177 E. Morava, R. A. Wevers, V. Cantagrel, L. H. Hoefsloot, L. Al-Gazali, J.
Schoots, A. van Rooij, K. Huijben, C. M. van Ravenswaaij-Arts, M. C.
Jongmans, J. Sykut-Cegielska, G. F. Hoffmann, P. Bluemel, M. Adamowicz,
J. van Reeuwijk, B. G. Ng, J. E. Bergman, H. van Bokhoven, C. Korner, D.
Babovic-Vuksanovic, M. A. Willemsen, J. G. Gleeson, L. Lehle, A. P. de
Brouwer and D. J. Lefeber, Brain, 2010, 133, 3210–3220.
178 E. Morava, R. Zeevaert, E. Korsch, K. Huijben, S. Wopereis, G. Matthijs, K.
Keymolen, D. J. Lefeber, L. De Meirleir and R. A. Wevers, Eur J Hum Genet,
2007, 15, 638–645.
179 B. G. Ng, C. Kranz, E. E. Hagebeuk, M. Duran, N. G. Abeling, B. Wuyts, D.
Ungar, V. Lupashin, C. M. Hartdorff, B. T. Poll-The and H. H. Freeze, Mol
Genet Metab, 2007, 91, 201–204.
180 B. G. Ng, V. Sharma, L. Sun, E. Loh, W. Hong, S. K. Tay and H. H. Freeze,
Mol Genet Metab, 2010, 102, 364–367.
154 | Carbohydr. Chem., 2012, 38, 124–155

181 J. Lubbehusen, C. Thiel, N. Rind, D. Ungar, B. H. Prinsen, T. J. de Koning, P.
M. van Hasselt and C. Korner, Hum Mol Genet, 19, 3623–3633.
182 E. Morava, S. Wopereis, P. Coucke, G. Gillessen-Kaesbach, T. Voit, J.
Smeitink, R. Wevers and S. Grunewald, Eur J Hum Genet, 2005, 13, 414–421.
183 M. Guillard, A. Dimopoulou, B. Fischer, E. Morava, D. J. Lefeber,
U. Kornak and R. A. Wevers, Biochim Biophys Acta, 2009, 1792, 903–914.
184 C. Noordam, S. Funke, N. V. Knoers, P. Jira, R. A. Wevers, Z. Urban and E.
Morava, Acta Paediatr, 2009, 98, 490–494.
Carbohydr. Chem., 2012, 38, 124–155 | 155

Bladder cancer–glycosylation insights
Paulo F. Severino,a,c Mariana Silva,a Mylène A. Carrascal,a
Fernando Calais,b Fabio Dall’Olioc and Paula A. Videira*a
DOI: 10.1039/9781849734769-00156
Bladder cancer is a common urologic cancer in Europe, with the highest recurrence
rate of any malignancy. Early diagnosis and correct staging of the disease is critical,
not only for planning treatment but also for selecting follow-up schedules, due to the
high risk of recurrence. Since few biomolecules have reached the clinical practice, an
unmet clinical need for biomarkers persists.
Correlations between specific glycan antigens and bladder cancer progression
stages, as well as diagnostic approaches based on glycan analysis have been sug-
gested worldwide. This chapter highlights the relevance and key aspects of the gly-
cosylation changes associated with bladder cancer. Moreover, recent studies suggest
that glycan-based markers would have widespread applicability in managing bladder
cancer treatment and are promising potential targets for novel therapies.
1 Introduction
Bladder cancer has one of the highest incidence rates in Europe, being the
fourth most common in men and the ninth in women. Non-muscle invasive
bladder cancer (NMIBC), the most frequent type, has a high recurrence rate
and represents the most costly common malignancy to manage per patient
from diagnosis to death.1–4
Bladder cancer signs and symptoms include haematuria, abdominal pain,
urinary urgency, pain while urinating, and loss of appetite. The risk factors
for the development of the disease are genetic and molecular abnormalities,
chronic irritation and infection, and chemical or environmental exposures.5–7
Cigarette smoking is the most common and also the main preventable risk
factor.8–10 Staging and grading of bladder tumours are carried out mainly
based on the cystoscopic and histopathologic analysis.11,12 However, the
current methodologies show poor sensitivity in detecting low grade disease,
and sometimes miss prediction of the long-term outcome of tumour lesions.13
While some biomarkers are becoming available to aid in the diagnosis
and staging, robust and specific biomarkers are still lacking. Besides their
usefulness in diagnosis, biomarkers could be also considered as potential
therapeutical targets.
One of the phenotypic changes associated with bladder cancer is the
expression of an aberrant glycosylation profile. In this chapter, we will
review the impact of altered glycosylation in bladder cancer. Consistent
with the book’s aim on Carbohydrates in Europe, in this chapter we will
highlight the latest European findings on bladder cancer glycome.
a
CEDOC, Departamento de Imunologia, Faculdade de Cieˆncias Médicas, FCM, Universidade
Nova de Lisboa, Lisboa, Portugal. E-mail: paula.videira@fcm.unl.pt
b
Grupo Português Geńito-Urinário, Hospital São Jose´, Lisboa, Portugal.
c
Dipartimento di Patologia Sperimentale, Università di Bologna, Bologna, Italia.
156 | Carbohydr. Chem., 2012, 38, 156–175

2 Bladder cancer: epidemiology and current treatment
Bladder cancer is a common malignancy, particularly in southern and
western European Countries (more than 30/100 000).12 Over 90% of blad-
der cancer cases are originated in the urothelial epithelial cells, being
referred as transitional cell carcinoma (TCC), approximately 5% are
squamous cell carcinoma, and less than 2% are adenocarcinoma. At diag-
nosis, more than 75% of TCC cases are non-muscle invasive (NMIBC),
including those that are superficial papillary tumours (Ta), invasive tumour
of the subepithelial connective tissue (T1), and carcinoma in situ (CIS). The
invasive forms, which are tumours that invade the muscle (T2), the peri-
vesical tissue (T3) and surrounding organs (T4) are believed to derive from
NMIBC.14
A decreasing trend in mortality has been evident in most European
countries over the past 10 years15 and presently, bladder cancer is generally
considered to have minimal risk of progression to death. However, about
60–70% of the patients develop recurrences, representing the highest
recurrence rate among solid tumours and 11% of these recurrences show
progression to muscle invasion. Most patients survive but require a lifelong
follow-up since more than half of the cases constantly relapse within one
year, requiring repeated courses of treatment.1,15–17
The treatment strategy of bladder cancer depends largely on its invasive
status. While invasive forms are usually treated by cystectomy, NMIBC are
treated by transurethral resection (TUR).6,12 The goal of treatment is two-
fold: to reduce tumour recurrence and subsequent need for supplementary
therapies and to prevent tumour progression and subsequent need for more
aggressive therapy. Several adjuvant agents, such as the Bacillus Calmette-
Guérin (BCG, an attenuated strain of Mycobacterium bovis), or che-
motherapeutic agents are essential as complementary therapies to reduce
risk for recurrence and progression of NMIBC.18 In patients with inter-
mediate to high risk for recurrence, BCG therapy is more effective than
other agents.12,14 Although the precise mechanisms responsible for BCG
anti-tumour effect have not been fully elucidated, a massive, complex and
local activation of the immune system has been well documented.19,20 After
intravesical instillation, BCG is internalised and processed both by pro-
fessional antigen-presenting cells (dendritic cells and macrophages) and
tumour cells. Then, BCG antigens trigger an inflammatory and immune
response which targets tumour cells.21 Recent in vitro findings support
the notion that the fibronectin attachment protein (FAP) from BCG has
an important role in directing BCG into cells.22,23 The activation of
the immune system results predominantly in a local synthesis of pro-
inflammatory cytokines and in the recruitment of several leukocyte
populations. Owing to its anatomy, the bladder offers advantages to adju-
vant therapy because it allows high local concentrations of either BCG or
any chemotherapeutic drug.21
However, 30–50% of the patients either fail to respond to BCG or recur
within the first year after treatment and in 10% of these patients the disease
progresses to muscle invasive and metastatic stages. Until now, there are no
biomarkers that can predict the patient’s prognosis and discriminate those
Carbohydr. Chem., 2012, 38, 156–175 | 157

that will respond to BCG treatment from those who would be best served by
more aggressive therapy such as cystectomy.
The diagnosis of bladder cancer depends mainly on cystoscopic exam-
ination of the bladder and histological evaluation of multiple tissue biop-
sies, and also includes urine cytology.12,24 As a standard procedure,
cystoscopy and TUR are performed using white light. However, since it can
miss lesions that are present but not visible, photodynamic diagnosis
(PDD), which uses violet light after intravesical instillation of 5-amino-
levulinic acid or hexaminolaevulinic acid, is also used. It has been confirmed
that fluorescence-guided biopsy and resection are more sensitive than con-
ventional procedures for detection of malignant tumours, particularly for
CIS. The additional detection rate with PDD was 20% for all tumours and
23% for CIS in a cumulative analysis of prospective trials.25
Like in any other cancers, screening of bladder cancer patients for tumour
specific biomarkers has been a principal goal of several research groups.26
The assessment of such biomarkers is of extreme importance to establish
non-invasive alternative diagnostic strategies, as well as reliable surveillance
markers. Several bladder cancer biomarkers have been described, such
as the mutated p53, homeobox protein HOX-1.3, keratins 8 and 13, psor-
iasin, adipocyte-type fatty acid-binding protein, GST Mu, 15-hydroxy-
prostaglandin dehydrogenase, tropomyosin 4, MRP-14, CD24, cytochrome
oxidase III subunit, the Nuclear Matrix Protein 22 (NMP-22) among oth-
ers, whose expression strongly correlates with a particular stage of bladder
cancer progression.27–38 Based on these biomarkers, different tissue and
urine test kits have been developed, whose sensibilities and specificities have
been reviewed by Van Tilborg et al., in 2009.17 Presently, urine tests such as
the bladder tumour antigen (BTA) stat test, which measures complement
factor H related protein is used in clinical practise.39 However, it is
becoming increasingly clear that no single marker will have the sensitivity
and specificity necessary to be used on its own for diagnosis/prognosis of
tumours. Interpatient and intratumour heterogeneity provides over-
whelming odds against the existence of such an ideal marker. Still, the use of
multiple biomarkers can be used for an improvement of the clinical
assessment of the disease.40
3 General aspects of glycosylation

Glycosylation is one of the most important modifications of proteins
and lipids and the majority glycans exist as membrane-bound or soluble
glycoconjugates. In eukaryotic cells, glycoconjugates are involved in a
variety of important biological mechanisms including cell adhesion, cell-cell
interaction, receptor activation and signal transduction.41 The three main
classes of glycoconjugates are glycoproteins, proteoglycans and glycolipids
(Fig. 1) and their synthesis occurs mainly in the lumen of the endoplasmic
reticulum and in the Golgi apparatus.
The sugar chains of glycoproteins are classified as N- or O-linked. The
N-linked chains are attached through the N-acetylglucosamine (GlcNAc)
residue to asparagine residue of the sequence N-X-S/T (where X is any
aminoacid, except proline), while the O-linked chains are usually linked
158 | Carbohydr. Chem., 2012, 38, 156–175

Fig. 1 Glycoconjugates in mammalian cells. Glycans can be attached to proteins as in
glycoproteins and proteoglycans. Glycoproteins are decorated with N- and/or O-glycans and
proteoglycan with glycosaminoglycans (GAG). Glycosphingolipids are mainly associated with
membranes and are formed by a glycan attached to the lipid moiety ceramide. A ganglioside
contains one or more residues of sialic acid (Sia). The glycan structures presented for each
N-linked, O-linked, glycoaminoglycan (GAG) and glycosphingolipids exemplify one of the
possible structures.
through an acetylgalactosamine (GalNAc) residue to serine or threonine

(Fig. 1). The biosynthesis of N-linked chains involves the en bloc transfer of
a preformed oligosaccharide comprised of two GlcNAc, nine mannoses
(Man) and three glucose (Glc) residues to the nascent polypeptide chain.42
Mature N-linked glycans consist of a core structure containing 2 GlcNAc
and 3 Man residues and a variable (2–4) number of antennas, usually
formed by GlcNAc, Galactose (Gal) and sialic acid (Sia). O-linked glycans
are attached to the hydroxyl group of serine or threonine on a protein. After
the addition of the first residue, the addition of other sugars is a stepwise
process which accompanies the maturation of the glycoprotein.43 In the
‘mucin-type’ O-glycans, the first GalNAc is further extended with Gal,
GlcNAc, and Sia (Fig. 1). There are also several types of non-mucin O-
glycans, including O-fucose (Fuc), O-xylose (Xyl), O-Man and O-GlcNAc
glycans.44 In this chapter, however, the term O-glycan refers to mucin
O-glycans.
The basic structure of proteoglycans is comprised of a core protein and
one or more covalently attached glycosaminoglycan (GAG) side chain.
GAGs are polysaccharides composed of repeating disaccharide units
formed by an amino sugar (GlcNAc in heparan sulphate or GalNAc in
chondroitin sulphate) linked to D-glucuronic acid (GlcA) or L-iduronic
acid (IdoA). The linkage of GAGs to the protein core involves a specific
Carbohydr. Chem., 2012, 38, 156–175 | 159

trisaccharide composed of two Gal and one Xyl sugars (GAG-GalGalXyl-
O-CH2-protein). GAGs exist also as free molecules, as is the case of hya-
luronan.45 Negatively charged sulphate groups of the heparan-sulphate
chains modulate interactions with different proteins involved in the cell–
matrix assemblage, inflammation and cellular division.45–47 In particular,
GAGs could play an important role in the tumour environment by facil-
itating the storage of important ligands.46
In glycolipids, the sugar portion is usually attached through a Glc residue
to the hydrophilic portion of a membrane lipid, which is often ceramide. In
this case the glycolipid is referred to as glycosphingolipid. Based on their
basic glycan structures, glycosphingolipids are classified into four groups
according to the sugar types linked to the Gal, namely, ganglio-series,
globo-series, paraglobo-series (or neolactoseries) and lacto-series. The gly-
cosphingolipid biosynthesis begins with the enzymatic addition of a Glc to
ceramide, forming glucosylceramide, which is then converted into lacto-
sylceramide by the addition of Gal. The successive addition of other sugar
residues, mediated by specific glycosyltransferases gives rise to the different
series of glycosphingolipids. The glycolipids of the ganglio-series, or gang-
liosides, may have one, two or three Sia residues linked to the inner Gal unit
and are thus termed a-, b- and c-series gangliosides, respectively (Fig. 2).
They include GM3, GM2, GM1, GD3 and GD2 and are characterized by
the presence of one or more Sia residues.48 Gangliosides are frequently
involved in various physiological and pathological processes, including cell
differentiation, migration, intracellular signalling, host-pathogen interac-
tions, and tumourigenesis.49–51
Unlike the biosynthesis of nucleic acids and proteins, which are deter-
ministic template-driven processes, glycosylation is a stochastic process,
regulated mainly by the relative abundance, cellular localization and spe-
cificities of the biosynthetic enzymes (glycosyltransferases) and catabolic
enzymes (glycosidases). The stochastic nature of the glycosylation process is
at the basis of the phenomenon known as microheterogeneity, which means
that the structure of the sugar chains attached to a specific glycosylation site
in a given glycoprotein displays a certain degree of variability.52–55
The expression of a given glycan structure which is usually cell type-
dependent is profoundly modified in cancer.44 In fact, an aberrant glyco-
sylation of proteins and lipids is frequently observed at the cancer cell
surfaces and can be shed into the body fluids of the patients (serum, urine,
pleural effusions, etc.). Aberrant glycosylation in tumour cells include both
a loss or a gain of expression of certain glycan structures, the appearance of
truncated structures, as well as of novel structures. Upregulation and/or
downregulation of specific glycosyltransferases is often responsible for these
changes.56–59 Tumour-associated carbohydrate structures allow tumour
cells to invade and metastasize.44 In the next sections, we will review some of
the most relevant glycosylation changes associated with bladder cancer.
3.1 Carbohydrate antigens in bladder cancer

3.1.1 Blood group antigens. Blood group antigens, which include the
ABH and Lewis blood group systems are a group of carbohydrate struc-
tures expressed at the terminal part of both glycoproteins (O-glycans and
160 | Carbohydr. Chem., 2012, 38, 156–175

A B
Fig. 2 Biosynthetic pathway of gangliosides and globosides. A: The conversion of lactosylceramide into GM3 and the formation of the main branches, the a- and
b-ganglioseries, are shown. B: Lactosylceramide may also be converted into Gb3 (also known as Pk) or paragloboside forming the two main branches, the globoside and
paragloboside series. Besides the structures, the enzymes responsible for the initial steps of their biosynthesis are also depicted.
Carbohydr. Chem., 2012, 38, 156–175 | 161

N-glycans) and glycolipids on red blood cells, endothelial cells and epithelial
cells, as well as in body fluids and secretions. The ABH and Lewis blood
group antigens are genetically and biochemically related, presenting a
common basic structure. These antigens are derived from substitutions on
type 1 (Galb1,3GlcNAc) or type 2 (Galb1,4GlcNAc) oligosaccharide
chains. The generation of the different glycan structures requires the con-
tribution of specific glycosyltransferases.60 The synthesis of the H antigen
consists in the transfer of a Fuc residue in a1,2 linkage to the Gal residue of
type 1 or 2 oligosaccharide chains (Table 1). The A or B antigens are
subsequently formed by the addition of a GalNAc or a Gal residue in
a1,3 linkage to the nonreducing terminus of the H antigen. The addition
Table 1 Carbohydrate sequence of ABH and Lewis blood antigens.
Type 1 oligossacharide chain
Type 2 oligossacharide chain
H antigen
A antigen
B antigen
Lewisa antigen
Lewisb antigen
Lewisx antigen
Lewisy antigen
sialyl Lewisa antigen
sialyl Lewisx antigen
162 | Carbohydr. Chem., 2012, 38, 156–175

of a Fuc residue in a1,4 linkage to the GlcNAc unit of type 1 chains
results in the biosynthesis of Lewisa (Lea) antigen, while the addition of Fuc
residue in a1,3 linkage to type 2 chains results in the biosynthesis of Lewisx
(Lex) antigen. If the addition of these two sugar residues occurs on the H
antigen, either the Lewisb (Leb) or the Lewisy (Ley) antigens are synthesized
(Table 1). If the addition of a1,3/4-linked Fuc is preceded by a2,3-sialyla-
tion of type 1 or of type 2 chains, the sialyl Lewisa (sLea) and sialyl Lewisx
(sLex) antigens, respectively, are synthesized.60
The expression of Lewis antigens is associated with the secretion of the A,
B or H antigens. In fact, the a1,2 fucosyltransferase expressed in secretory
glands (FucT2) is responsible for both, Lea to Leb conversion and the
addition of fucose to A, B and H antigens on soluble molecules. Therefore,
people with Lea antigens will not secrete the A, B or H antigens (ABH non-
secretors), and the presence of the Leb antigen is only found in secretors.
The altered expression of blood group antigens is a common feature of
many epithelial cancers. The loss of the expression of blood group antigens
has been observed in all different tumour stages of TCC and it has been
associated with tumour progression, recurrent disease and metastases for-
mation.61,62 Immunohistochemical analyses revealed that the Lex antigen is
expressed in TCC, regardless the secretory status of the individual and the
stage of the tumour,61,63 while the expression of sLex was found to be
related with invasive and metastatic potential.64 Several publications
reported that the expression of sLea antigen was upregulated in bladder
cancer when compared with other urologic diseases.65–69 Moreover, its
expression has been recently correlated with tumour invasion and grade.70
Accordingly, the expression of either sLex or sLea is known to allow cancer
cells to evade the immune responses and mimic the ability of blood cells to
migrate and spread through binding to endothelial selectins.50,71 Recently, it
was reported that Ley antigen expression was also altered in bladder cancer,
with different expression profiles between CIS and non-CIS condition of the
bladder.72 Since CIS lesions are associated with high-risk of progression,
but no accurate biomarkers have ever been established, the inclusion of new
glycans such as Ley would be of help in the establishment of a glycoprofile
for this type of lesions.
The loss of ABH structures and the increased expression of Lewis anti-
gens are the most evident changes, among the blood group antigens, related
with poor prognosis in bladder cancer,70 suggesting a pathophysiological
role in bladder cancer progression.
3.1.2 Thomsen-Friedenreich related antigens. Thomsen-Friedenreich
(TF) related antigens are a group of O-glycans which includes the Tn and T
antigens, and their sialylated forms sialyl Tn (sTn) and sialyl T (sT) antigens
(Fig. 3). The Tn antigen (GalNAc-O-Ser/Thr) is the precursor of the T
antigen (Gala1,3GalNAc-O-Ser/Thr), also known as core 1 structure. sTn is
produced by the addition of Sia in a2,6-linkage to GalNAc. The bio-
synthesis of sT results mainly from the a2,3-sialylation of the Gal residue of
T antigen43 (3-sialyl T antigen, Fig. 3), but it also includes the a2,6-sialy-
lation of the GalNAc (disialyl T antigen) or the Gal residue (6-sialyl T
antigen). Sialylation of TF-related glycans blocks the elongation of the
Carbohydr. Chem., 2012, 38, 156–175 | 163

Fig. 3 Structures of Thomsen-Friedenreich related antigens. Both glycan structures and the
main enzymes responsible for their synthesis are shown.
sugar chain, resulting in the accumulation of these glycans which, in turn,

affect the adhesive properties of cancer cells, their invasive and angiogenic
potential as well as their recognition by immune cells.51,73
In bladder cancer, the presence and the clinical importance of TF-related
antigens is still controversial.74–78 The earliest evidence of a deregulated
expression of TF-related antigens in bladder cancer came from the obser-
vation of an enhanced binding of the peanut lectin (PNA), which recognizes
the T antigen, to the bladder tumour tissue.79,80 The PNA staining is a
largely used method for the detection of T antigens in human normal and
malignant tissues because it requires little amount of samples. The expres-
sion of sT antigens can be inferred after sialidase treatment and staining
with PNA. In bladder, it was also reported that the PNA binding was
correlated with invasion and a higher risk of lymph node metastasis.79,81
However, the aberrant expression of the sialylated forms of T was also
reported.76,77 The ‘‘cryptic’’ T is present in the healthy urothelium but
overexpressed in bladder tumors, in accordance with the author’s obser-
vations. A deeper understanding of the structural nature of these ‘‘cryptic’’
glycans may render in the future valuable information regarding bladder
cancer Glycobiology.
Besides the TF-related antigens, also the sialyltransferases involved in the
biosynthesis of their sialylated forms are potential prognostic markers. In
fact, when we analysed the mRNA expression level of the sialyltransferases
involved in the synthesis of the sTn and sT antigens in bladder cancer and
normal urothelium biopsies, we found that the expression of ST3Gal.I57
and of ST6GalNAc.I (unpublished data) were augmented in bladder cancer.
In the same study, a higher expression of ST3Gal.I and a decreased
expression of ST3Gal.II in tumour samples correlated with propensity to
tumour recurrence.57 Concordantly with the increased ST6GalNac.I
expression, the analysis of histological sections from bladder cancer patients
revealed the presence of sTn antigens apparently related with a low degree
of dedifferentiation of the malignant cells (Videira et al., unpublished data).
While a correlation has been shown in vitro between ST3Gal.I and
sT expression, the same was not observed in patient’s samples.57 Further
164 | Carbohydr. Chem., 2012, 38, 156–175

investigations, involving larger number of patients, are still needed to
identify the targets of the above mentioned sialyltransferases and the TF-
decorated glycoproteins.
A glycopeptide decorated with a trisaccharide similar, if not identical,
with sT was identified in the urine of patients affected by interstitial cystitis,
a painful disease caused by thinning or denudation of bladder epithelium.
This molecule, which is referred to as antiproliferative factor (APF) exerts a
strong anti-proliferative activity on bladder cells and is involved in the
pathogenesis of interstitial cystitis. The peptide backbone is related to
frizzled 8, a G protein-coupled receptor of the Wnt signalling pathway and
the presence of the sugar chain is necessary for the antiproliferative activity
of APF.82 Altogether, these observations are consistent with a role for
TF-related glycans in the control of bladder cell proliferation.
3.1.3 Glycosphingolipids. Several reports showed drastic changes in the

glycosylation of lipids in bladder tumours and correlated these modifica-
tions with tumour invasiveness and progression.83,84 While NMIBC show
an accumulation of GM3 gangliosides and a decrease of Gb3 and Gb4
globo-series structures, invasive tumours show a decreased expression of
GM3 and an accumulation of Gb3 and Gb4 structures.83 The accumulation
of GM3 structures in NMIBC may be explained by the upregulation of
GM3 synthase and the downregulation of Gb3 and Gb4 synthase.83
Moreover, it has been demonstrated that the exogenous addition of GM3
inhibits the invasive activity of bladder cancer cell lines. Concordantly,
bladder cancer cell lines overexpressing GM3 synthase have reduced pro-
liferation potential, motility and growth as xenografts.83,85 Todeschini et al.
reported that GM2/GM3 complexes efficiently inhibit cell motility of two
different bladder cancer cell lines.86 More recently, it was described the
involvement of specific glycosphingolipids, including GM3, GM2 and
GM1, in the epithelial-to-mesenchymal cell transition (EMT) process of
bladder cells, a basic process of cancer progression characterized by loss of
cell adhesion and increased cell mobility.84
3.2 Glycoproteins
3.2.1 Cadherins. The cadherin cell adhesion molecules are calcium-
dependent transmembrane glycoproteins and are the main constituent of
the adherens junctions. The cadherins are comprised of extracellular,
membrane and cytoplasmic domains with the cytoplasmic domain anchored
to the cell cytoskeleton by the catenin family members.87 Cadherins are
implicated in numerous signalling events related to embryonic development,
tissue morphogenesis and homeostasis and are the main mediators of cell-
cell adhesion in normal and cancer epithelial tissues.88–90 In several studies,
the downregulation of E-cadherin in malignant human urothelium was
associated with tumour recurrence, progression and survival.91–100 In vitro
studies in bladder cancer cell lines also suggested that low E-cadherin
expression was correlated with expansion of CIS.101–103 On the other hand,
higher levels of the proteolytic cleavage product of the E-cadherin, sE-
cadherin, can be found particularly in serum of patients with TCC, being
detectable also in urine.104–107 In addition, in bladder cancer it has also been
Carbohydr. Chem., 2012, 38, 156–175 | 165

observed a switch from the production of E-cadherin to that of N-cadherin
(typically found in neural and mesodermal tissues) and P-cadherin (typically
found in placental tissue). These events seem to be important late processes
in the progression of bladder cancer.108 Cadherins switching has also been
reported in several other malignancies.109 A study on non-malignant and
malignant bladder cancer cell lines has revealed a tendency to accumulate
b1,6-branched oligosaccharides with poly-N-acetyllactosaminic structures
and a2,3-linked Sia residues in the latter,110 which are usually associated
with metastasis.111 However, N-acetylglucosaminyltransferase V (GnT-V),
a key enzyme in the formation of the b1,6-branched N-linked oligo-
saccharides, was inversely correlated with recurrence events and directly
correlated with low malignant potential and good prognosis.112 Further
studies seem to be necessary to understand the pathological role of b1,6-
branched N-linked glycans on E-cadherins.
3.2.2 Integrins. Integrins are a widely distributed family of hetero-
dimeric cell membrane glycoproteins involved in cell–to–extracellular
matrix and cell–to–cell interactions, and mediating important intracellular
signalling cascades.113,114 These glycoproteins consist of sixteen types of a
subunits and eight types of b subunits that assemble into twenty two types
of cell surface receptors that bind mostly large molecules, usually found in
the subendothelial matrix, including fibronectin, vitronectin and col-
lagen.113,115–118 Integrins are involved in cell growth and differentiation,
proliferation and migration, tissue organization, recruiting of lymphocytes
and inflammation, cancer invasion and metastasis.113 Integrin expression
and glycosylation are altered upon cell transformation, resulting in altered
function of these molecules. In the bladder cancer cells, but not in normal
epithelium, it was found the presence of a sialylated tetraantennary N-
oligosaccharide in both subunits of a3b1 integrin.119 An aberrant expres-
sion of b1-6 branched N-linked glycans, as well as high mannose type gly-
cans, has also been detected in a3b1 integrins from bladder cancer cells.
Removal of Sia and treatment with swainsonine (an inhibitor of Golgi a-
mannosidase II, an important enzyme in N-glycan maturation) in T24
bladder cancer cells resulted in a stronger fibronectin-binding activity of
a3b1 integrin and in a lower rate of tumour cells migration respectively,
suggesting that the glycosylation profile of a3b1 integrin may have an
important function in bladder cancer cells migration.119
3.2.3 Mucins. Mucins (MUCs) are heavily glycosylated high-molecular-
mass glycoproteins that can be present both in membrane-bound and
secreted forms and represent a major component of mucus.120 The abun-
dance of serine and threonine residues in their variable number of tandem
repeats provides the backbone necessary for their bulk O-glycosylation.121
In TCC (transitional cell carcinoma), the pattern, intensity and depth of
MUC1 staining were significantly associated with tumour stage and grade
and metastases.122 However, larger studies are required to define more
precisely the role of MUC1 in the prognosis of bladder cancer. The anti-
MUC1, C595, monoclonal antibody labelled with 111Indium or 67Cupper,
when intravenously administrated was able to detect bladder cancer cells,
including distant metastasis,123,124 suggesting the possible therapeutic use of
166 | Carbohydr. Chem., 2012, 38, 156–175

anti-MUC1 antibodies conjugated with cytotoxic isotopes. MUC7 is also
atypically expressed in bladder cancer, in the soluble form, being detected
into the urine of patients.125 The glycosylation of both MUC1 and MUC7
varies among different issue and in cancer they have been reported to be more
or less glycosylated than the corresponding normal tissues. Until now, the
glycosylation profile of both mucins remains unknown. The identification of
minute glycosylation changes in these mucins is likely to improve the diag-
nostic potential of MUC1 and 7 as bladder cancer biomarkers.
3.3 Extracellular matrix

Invasion by tumour cells involves the alteration of cell–matrix interac-
tions.51 Extracellular matrix is comprised of a complex mixture of macro-
molecules capable of self-assemble, composed predominantly of collagens,
non-collagenous glycoproteins, elastin, hyaluronan and proteoglycans. The
extracellular matrix represents not only a scaffold for the cells, but also a
reservoir for important biomolecules, such as growth factors and cytokines.
In addition, there is increasing evidence that extracellular matrix molecules
may exhibit a direct signalling function, either via interactions with matrix
receptors like integrins or via signalling through growth factor receptors
themselves.126,127 These biomolecules may also play important roles in
cancer growth.126 Hyaluronic acid (HA), a nonsulfated GAG comprised of
a repeating disaccharide (GlcA-GlcNAc) absorbs significant amounts of
water in the extracellular space and confers tissues the ability to resist
compression. The expression of HA has been implicated in promoting
tumour cell proliferation, invasion and metastasis and represents an accu-
rate diagnostic marker for bladder cancer. On the other hand, hyalur-
onidase (HAase) degrades HA into proangiogenic fragments that support
tumour metastasis. The measurement of HA and of its degrading enzyme
hyaluronidase (HAase) in urine has been applied as a commercial urinary
test, the HA-HAase test, to the screening of bladder cancer.128,129
3.4 Carbohydrate-binding proteins – Lectins

Initial insights into the unique repertoire of glycans expressed on tumour
cells emerged from the increased ability of tumours to bind a range of plant
lectins.130 In fact, a large number of the studies on the alterations of gly-
cosylation in cancer has been performed using lectins as probes.131 Lectins
exhibit a typical protein folding that defines the carbohydrate-binding
protein families and bind defined monosaccharide or oligosaccharide
structures with good specificity131 in a ‘lock-and-key’ fashion.51,132 The
carbohydrate-binding proteins are divided into several types depending on
their structures and the nature of the interaction they establish.131 Long
before the role of carbohydrate–protein interactions had been explored,
many lectins were identified, characterized and applied as useful tools in
studying glycoconjugates. Therefore, lectins have been widely used for
preparative and analytical purposes in biochemistry, cell biology, immu-
nology and related areas.131
Some lectins were found to be useful in differentiating bladder tumours
from normal urothelium. Lectins like Ricinus communis lectin (RCA,
recognizing Galb1,4GlcNAcb1-R) and wheat germ lectin (WGA,
Carbohydr. Chem., 2012, 38, 156–175 | 167

recognizing GlcNAcb1,4GlcNAcb1,4GlcNAc-R) can bind to all urothelial
cells in the normal tissue, while some lectins like PNA, concanavalin A
lectin (ConA, recognizing branched a-mannosidic structures and bian-
tennary complex type N-Glycans) and soybean lectin (SBA, recognizing
GalNAc) show little binding.133 Initially, by using cell suspensions from
transitional cell carcinoma, it was found that aneuploid cells bound PNA
more extensively and were less reactive with WGA than diploid cell popu-
lation.77 Drugs or delivery systems modified by lectins have been thought as
potential tools for more efficient, local adjuvant intravesical treatment for
bladder cancer.134
Galectins represent a family of calcium-independent Gal-specific lectins,
widely distributed in all living organisms.135 There are 15 family members
reported in humans, each of which plays a distinct role in adhesion, dif-
ferentiation, growth, apoptosis, etc.136–139 Galectins play a number of
important roles in cancer by contributing to tumourigenesis, proliferation,
angiogenesis, and metastasis.140–145 These lectins are the best characterized
carbohydrate-binding proteins in bladder cancer.136,139,146–148 The expres-
sion of galectin-1 and galectin-3 was investigated in human bladder tran-
sitional cell carcinomas of different histological grade and clinical stage and
in normal urothelium samples.145 Galectin-1 levels were found to be highly
increased in most high-grade tumours when compared with normal bladder
or low-grade tumours. Galectin-3 mRNA levels were also found increased
in most tumours, however with no relevant differences among tumours of
different histological grade.145 In a more detailed study, it was found an
increased expression of the galectin-3 mRNA in invasive bladder tumours
when compared with NMIBC and an association with poor survival; while
galectin-3 protein levels in urine allowed identifying bladder cancer
patients.149 Galectin-3 plays an important role in bladder cancer by inhi-
biting TNF-related apoptosis-inducing ligand (TRAIL)-induced apopto-
sis.150 Furthermore, the development of anti-galectin compounds has been
suggested as a potential target for therapeutic intervention in bladder can-
cer.151 In addition galectins-1 and -3 have also been considered bladder
tumour markers.145 The expression of galectin-7 had been hypothesised to
be related to the sensitivity of bladder cancer to platinum-containing drugs.
In fact, galectin-7 levels, which were higher in normal urothelium than in
bladder cancer, were found to increase in response to treatment with pla-
tinum-containing drugs, but only in p53 wild type tumours. Moreover,
transfection of galectin-7 cDNA in p53-mutated cell lines resulted in
increased drug sensitivity.152
4 Final considerations/conclusions
Owing to its high incidence and percentage of recurrence, bladder cancer
represents a major public health concern. Consequently, any effort should
be made to improve the survival and the quality of life of the patients. In
this light, the development of strategies aimed at early diagnosis and at
more effective therapies represents major goals.
Along this chapter, we have provided an overview of how Glycobiology
can contribute to pursue these achievements. We have described the main
168 | Carbohydr. Chem., 2012, 38, 156–175

findings of basic and clinical research, showing how sugar chains, glyco-
sylated molecules and sugar binding molecules can affect the progression of
this malignancy. As biomarker discovery becomes more sophisticated,
glycobiology is being used to bring more sensitive and specific diagnostic
based on cancer-related changes in glycosylation on specific glycoproteins.
We are confident that research in the field of glycobiology will offer in the
future new diagnostic and therapeutic approaches with a strong positive
impact on the life of the patients.
Abbreviations
APF antiproliferative factor
BCG Bacillus Calmette-Gue´rin
CIS carcinoma in situ
Fuc fucose
GAG glycosaminoglycan
Gal galactose
GalNac N-acetylgalactosamine
Glc glucose
GlcA D-glucuronic acid
GlcNAc N-acetylglucosamine
HA hyaluronic acid
HAase hyaluronidase
IdoA L-iduronic acid
Lea Lewisa
Leb Lewisb
Lex Lewisx
Ley Lewisy
Man mannose
MUC mucin
NMIBC non-muscle invasive bladder cancer
PDD photodynamic diagnosis
PNA peanut lectin
Sia sialic acid
sLea sialyl Lewisa
sLex sialyl Lewisx
sT sialyl T
sTn sialyl Tn
TCC transitional cell carcinoma
TF Thomsen-Friedenreich-related antigens
TUR transurethral resection
References
1 M. L. Hemani and C. L. Bennett, Cancer, 2010, 116, 3530–3532.
2 T. P. Kresowik and T. S. Griffith, Immunotherapy, 2009, 1, 281–288.
3 M. Racioppi, A. Volpe, R. Falabella, F. Pinto, E. Sacco, G. Gulino, D.
D’Agostino and P. F. Bassi, Arch Ital Urol Androl, 2007, 79, 111–117.
4 A. Stenzl, J. Hennenlotter and D. Schilling, Curr Opin Urol, 2008, 18, 488–492.
Carbohydr. Chem., 2012, 38, 156–175 | 169

5 C. Pelucchi, C. Bosetti, E. Negri, M. Malvezzi and C. La Vecchia, Nat Clin
Pract Urol, 2006, 3, 327–340.
6 D. S. Kaufman, W. U. Shipley and A. S. Feldman, Lancet, 2009, 374, 239–249.
7 M. J. Thun, S. J. Henley and T. Gansler, Novartis Found Symp, 2004, 256,
6–21; discussion, 22-28, 49-52, 266–269.
8 B. Anderson, Br J Community Nurs, 2009, 14, 385–392.
9 M. P. Zeegers, F. E. Tan, E. Dorant and P. A. van Den Brandt, Cancer, 2000,
89, 630–639.
10 P. Boffetta, Scand J Urol Nephrol Suppl, 2008, 45–54.
11 M. Burchardt, T. Burchardt, A. Shabsigh, A. De La Taille, M. C. Benson and
I. Sawczuk, Clin Chem, 2000, 46, 595–605.
12 M. Babjuk, W. Oosterlinck, R. Sylvester, E. Kaasinen, A. Böhle, J. Palou-
Redorta and M. Rouprêt, and E. A. o. U. (EAU), Eur Urol, 2011, 59, 997–
1008.
13 D. Mitropoulos, A. Kiroudi-Voulgari, P. Nikolopoulos, T. Manousakas and
A. Zervas, J Endourol, 2005, 19, 861–864.
14 B. L. Jacobs, C. T. Lee and J. E. Montie, CA Cancer J Clin, 2010, 60, 244–272.
15 D. M. Parkin, Scand J Urol Nephrol Suppl, 2008, 12–20.
16 D. M. Latini, S. P. Lerner, S. W. Wade, D. W. Lee and D. Z. Quale, Urology,
2010, 75, 334–339.
17 A. A. Van Tilborg, C. H. Bangma and E. C. Zwarthoff, Int J Urol, 2009, 16,
23–30.
18 K. Hendricksen and J. A. Witjes, Curr Opin Urol, 2007, 17, 352–357.
19 A. Bohle and S. Brandau, The Journal of Urology, 2003, 170, 964–969.
20 P. A. Videira, F. M. Calais, M. Correia, D. Ligeiro, H. J. Crespo, F. Calais
and H. Trindade, Urology, 2009.
21 R. F. Bevers, K. H. Kurth and D. H. Schamhart, British Journal of Cancer,
2004, 91, 607–612.
22 B. G. Coon, S. Crist, A. M. Gonzalez-Bonet, H. K. Kim, J. Sowa, D. H.
Thompson, T. L. Ratliff and R. C. Aguilar, Int J Cancer, 2011.
23 H. W. Sinn, B. D. Elzey, R. J. Jensen, X. Zhao, W. Zhao and T. L. Ratliff,
Cancer Immunol Immunother, 2008, 57, 573–579.
24 K. H. Kurth, P. F. Schellhammer, E. Okajima, A. Akdas, G. Jakse, H. W.
Herr, F. Calais da Silva, S. Fukushima and T. Nagayama, Int J Urol., 1995,
2(Suppl 2), 8–22.
25 J. A. Witjes, P. M. Moonen and A. G. van der Heijden, European Urology,
2005, 47, 319–322.
26 D. Tilki, M. Burger, G. Dalbagni, H. B. Grossman, O. W. Hakenberg, J.
Palou, O. Reich, M. Rouprêt, S. F. Shariat and A. R. Zlotta, Eur Urol, 2011,
60, 484–492.
27 J. E. Celis, M. Ostergaard, B. Basse, A. Celis, J. B. Lauridsen, G. P. Ratz, I.
Andersen, B. Hein, H. Wolf, T. F. Orntoft and H. H. Rasmussen, Cancer Res,
1996, 56, 4782–4790.
28 I. Gromova, P. Gromov and J. E. Celis, Electrophoresis, 1999, 20, 241–248.
29 M. Ostergaard, H. Wolf, T. F. Orntoft and J. E. Celis, Electrophoresis, 1999,
20, 349–354.
30 I. Gromova, P. Gromov, H. Wolf and J. E. Celis, Int J Oncol, 1998, 13,
379–383.
31 J. E. Celis, P. Celis, H. Palsdottir, M. Ostergaard, P. Gromov, H. Primdahl, T.
F. Orntoft, H. Wolf, A. Celis and I. Gromova, Mol Cell Proteomics, 2002, 1,
269–279.
32 M. Ostergaard, H. H. Rasmussen, H. V. Nielsen, H. Vorum, T. F. Orntoft, H.
Wolf and J. E. Celis, Cancer Res, 1997, 57, 4111–4117.
170 | Carbohydr. Chem., 2012, 38, 156–175

33 G. Ohlsson, J. M. Moreira, P. Gromov, G. Sauter and J. E. Celis, Mol Cell
Proteomics, 2005, 4, 570–581.
34 J. M. Moreira, P. Gromov and J. E. Celis, Mol Cell Proteomics, 2004, 3,
410–419.
35 A. Saad, D. C. Hanbury, T. A. McNicholas, G. B. Boustead, S. Morgan and
A. C. Woodman, BJU Int, 2002, 89, 369–373.
36 V. Poulakis, U. Witzsch, R. De Vries, H. M. Altmannsberger, M. J. Manyak
and E. Becht, BJU Int, 2001, 88, 692–701.
37 K. H. Tsui, S. M. Chen, T. M. Wang, H. H. Juang, C. L. Chen, G. H. Sun and
P. L. Chang, Asian J Androl, 2007, 9, 711–715.
38 P. Bassi, V. De Marco, A. De Lisa, M. Mancini, F. Pinto, R. Bertoloni and
F. Longo, Urol Int, 2005, 75, 193–200.
39 R. Kinders, T. Jones, R. Root, C. Bruce, H. Murchison, M. Corey, L.
Williams, D. Enfield and G. M. Hass, Clin Cancer Res, 1998, 4, 2511–2520.
40 J. M. Moreira, G. Ohlsson, P. Gromov, R. Simon, G. Sauter, J. E. Celis and
I. Gromova, Mol Cell Proteomics, 2010, 9, 161–177.
41 A. Varki, Glycobiology, 1993, 3, 97–130.
42 R. Kornfeld and S. Kornfeld, Annu Rev Biochem, 1985, 54, 631–664.
43 I. Brockhausen, Biochim Biophys Acta, 1999, 1473, 67–95.
44 A. Varki, R. Kannagi and B. P. Toole, in Essentials of Glycobiology, eds. A.
Varki, R. D. Cummings, J. D. Esko, H. H. Freeze, P. Stanley, C. R. Bertozzi,
G. W. Hart and M. E. Etzler, The Consortium of Glycobiology Editors, La
Jolla, California, Cold Spring Harbor (NY) Editon edn., 2009, vol. 2nd.
45 J. D. Esko and U. Lindahl, J Clin Invest, 2001, 108, 169–173.
46 M. Bernfield, M. Gotte, P. W. Park, O. Reizes, M. L. Fitzgerald, J. Lincecum
and M. Zako, Annu Rev Biochem, 1999, 68, 729–777.
47 J. D. Esko, K. Kimata and U. Lindahl, in Essentials of Glycobiology, eds. A.
Jolla, California, Cold Spring Harbor NYEditon edn., 2009.
48 R. L. Schnaar, A. Suzuki and P. Stanley, in Essentials of Glycobiology, eds. A.
Jolla, California, Cold Spring Harbor NYEditon edn., 2009.
49 C. T. Campbell and K. J. Yarema, Genome biology, 2005, 6, 236.
50 C. A. Reis, H. Osorio, L. Silva, C. Gomes and L. David, Journal of clinical
pathology, 2010, 63, 322–329.
51 M. M. Fuster and J. D. Esko, Nat. Rev. Cancer., 2005, 5, 526–542.
52 G. Lauc and V. Zoldos, Mol Biosyst, 2010, 6, 2373–2379.
53 G. Lauc, I. Rudan, H. Campbell and P. M. Rudd, Mol Biosyst, 2010, 6,
329–335.
54 K. L. Abbott, A. V. Nairn, E. M. Hall, M. B. Horton, J. F. McDonald, K. W.
Moremen, D. M. Dinulescu and M. Pierce, Proteomics, 2008, 8, 3210–3220.
55 A. V. Nairn, W. S. York, K. Harris, E. M. Hall, J. M. Pierce and K. W.
Moremen, The Journal of Biological Chemistry, 2008, 283, 17298–17313.
56 P. A. Videira, I. F. Amado and M. Correia, F. M. Calais da Silva, F. Calais da
Silva, F. Dall’Olio and H. Trindade, URO, 2007, 13, 26–32.
57 P. A. Videira, M. Correia, N. Malagolini, H. J. Crespo and D. Ligeiro, F. M.
Calais, H. Trindade and F. Dall’Olio, BMC Cancer, 2009, 9, 357.
58 F. Dall’olio, N. Malagolini and M. Chiricolo, in Carbohydrate Chemistry, The
Royal Society of Chemistry, Editon edn., 2012.
59 F. Dall’olio, N. Malagolini, M. Trinchera and M. Chiricolo, Front Biosci,
2012, 17, 670–699.
Carbohydr. Chem., 2012, 38, 156–175 | 171

60 P. Stanley and R. D. Cummings, 2009.
61 J. Sheinfeld, V. E. Reuter, A. S. Sarkis and C. Cordon-Cardo, J Cell Biochem
Suppl, 1992, 16I, 50–55.
62 C. Cordon-Cardo, V. E. Reuter, K. O. Lloyd, J. Sheinfeld, W. R. Fair, L. J.
Old and M. R. Melamed, Cancer Res, 1988, 48, 4113–4120.
63 J. Sheinfeld, V. E. Reuter, W. R. Fair and C. Cordon-Cardo, Semin Surg
Oncol, 1992, 8, 308–315.
64 K. Numahata, M. Satoh, K. Handa, S. Saito, C. Ohyama, A. Ito, T. Taka-
hashi, S. Hoshi, S. Orikasa and S. I. Hakomori, Cancer, 2002, 94, 673–685.
65 A. Tizzani, G. Cassetta, A. Cicigoi, P. Piana, A. Cerchier, F. Pecchio and P.
Piantino, Int J Biol Markers, 1987, 2, 121–124.
66 O. A. Mansour, S. A. Shouman, M. Khafagy, H. Basma, P. M. Liebmann, K.
Schauenstein and O. el-Ahmady, Lancet, 1995, 346, 512–513.
67 K. Noto, M. Fujime, H. Isobe, Y. Wakumato and Y. Kawachi, Nihon
Hinyokika Gakkai Zasshi, 1997, 88, 406–413.
68 E. M. Vestergaard, H. Wolf and T. F. Orntoft, Clin Chem, 1998, 44, 197–204.
69 K. Nagao, Y. Itoh, K. Fujita and M. Fujime, Int J Urol, 2007, 14, 795–799.
70 A. Hegele, V. Mecklenburg, Z. Varga, P. Olbert, R. Hofmann and P. Barth,
Anticancer Res, 2010, 30, 5195–5200.
71 M. Ugorski and A. Laskowska, Acta Biochimica Polonica, 2002, 49, 303–311.
72 S. Gunia, C. Kakies, M. May, S. Koch and A. Erbersdobler, J Clin Pathol,
2011, 64, 672–676.
73 F. Dall’Olio and M. Chiricolo, Glycoconjugate Journal, 2001, 18, 841–850.
74 Z. F. Dobrowolski, D. Dus, J. Halasa and C. Radzikowski, International
Urology and Nephrology, 1995, 27, 395–404.
75 N. C. Langkilde, Scandinavian Journal of Urology and Nephrology, Supple-
mentum, 1995, 172, 45–49.
76 N. C. Langkilde and T. F. Orntoft, Scandinavian Journal of Urology and
Nephrology. Supplementum, 1995, 172, 57–64.
77 N. C. Langkilde, H. Wolf, H. Clausen and T. F. Orntoft, Cancer Research,
1992, 52, 5030–5036.
78 C. B. Pinnock, D. J. Roxby, J. M. Ross, C. H. Pozza and V. R. Marshall,
British Journal of Urology, 1995, 75, 461–469.
79 C. Limas and P. Lange, Cancer, 1986, 58, 1236–1245.
80 T. P. Lehman, H. S. Cooper and S. G. Mulholland, Cancer, 1984, 53, 272–277.
81 K. Nakanishi, T. Kawai and M. Suzuki, Histopathology, 1993, 23, 153–158.
82 S. K. Keay, Z. Szekely, T. P. Conrads, T. D. Veenstra, J. J. Barchi, Jr., C. O.
Zhang, K. R. Koch and C. J. Michejda, Proceedings of the National Academy
of Sciences of the United States of America, 2004, 101, 11803–11808.
83 S. Kawamura, C. Ohyama, R. Watanabe, M. Satoh, S. Saito, S. Hoshi, S.
Gasa and S. Orikasa, Int J Cancer, 2001, 94, 343–347.
84 F. Guan, K. Handa and S. I. Hakomori, Proc Natl Acad Sci U S A, 2009, 106,
7461–7466.
85 R. Watanabe, C. Ohyama, H. Aoki, T. Takahashi, M. Satoh, S. Saito,
S. Hoshi, A. Ishii, M. Saito and Y. Arai, Cancer Research, 2002, 62, 3850–
3854.
86 A. R. Todeschini, J. N. Dos Santos, K. Handa and S. I. Hakomori, Proc Natl
Acad Sci U S A, 2008, 105, 1925–1930.
87 F. Nollet, P. Kools and F. van Roy, J Mol Biol, 2000, 299, 551–572.
88 U. Cavallaro, B. Schaffhauser and G. Christofori, Cancer Lett, 2002, 176,
123–128.
89 P. Hulpiau and F. van Roy, Int J Biochem Cell Biol, 2009, 41, 349–369.
90 A. Jeanes, C. J. Gottardi and A. S. Yap, Oncogene, 2008, 27, 6920–6929.
172 | Carbohydr. Chem., 2012, 38, 156–175

91 A. Clairotte, I. Lascombe, S. Fauconnet, F. Mauny, S. Felix, M. P. Algros,
H. Bittard and B. Kantelip, Am J Clin Pathol, 2006, 125, 119–126.
92 P. P. Bringuier, R. Umbas, H. E. Schaafsma, H. F. Karthaus, F. M. Debruyne
and J. A. Schalken, Cancer Res, 1993, 53, 3241–3245.
93 R. R. Byrne, S. F. Shariat, R. Brown, M. W. Kattan, R. J. Morton, T. M.
Wheeler and S. P. Lerner, J Urol, 2001, 165, 1473–1479.
94 P. K. Lipponen and M. J. Eskelinen, J Cancer Res Clin Oncol, 1995, 121,
303–308.
95 A. Mialhe, J. Louis, S. Montlevier, M. Peoch, D. Pasquier, J. L. Bosson, J. J.
Rambeaud and D. Seigneurin, Invasion Metastasis, 1997, 17, 124–137.
96 J. S. Ross, A. D. del Rosario, H. L. Figge, C. Sheehan, H. A. Fisher and H. X.
Bui, Hum Pathol, 1995, 26, 940–944.
97 R. L. Hong, Y. S. Pu, T. S. Hsieh, J. S. Chu and W. J. Lee, J Urol, 1995, 153,
2025–2028.
98 K. N. Syrigos, K. Harrington, J. Waxman, T. Krausz and M. Pignatelli,
J Urol, 1998, 160, 1889–1893.
99 T. Otto, A. Bex, U. Schmidt, A. Raz and H. Rubben, Am J Pathol, 1997, 150,
1919–1923.
100 Z. Popov, S. Gil-Diez de Medina, M. A. Lefrere-Belda, A. Hoznek, S. Bastuji-
Garin, C. C. Abbou, J. P. Thiery, F. Radvanyi and D. K. Chopin, Br J Cancer,
2000, 83, 209–214.
101 E. M. Bindels, M. Vermey, R. van den Beemd, W. N. Dinjens and T. H. Van
Der Kwast, Cancer Res, 2000, 60, 177–183.
102 U. H. Frixen, J. Behrens, M. Sachs, G. Eberle, B. Voss, A. Warda, D. Löchner
and W. Birchmeier, The Journal of Cell Biology, 1991, 113, 173–185.
103 K. Vleminckx, L. Vakaet, Jr., M. Mareel, W. Fiers and F. van Roy, Cell, 1991,
66, 107–119.
104 S. F. Shariat, K. Matsumoto, R. Casella, W. Jian and S. P. Lerner, Eur Urol,
2005, 48, 69–76.
105 K. Matsumoto, S. F. Shariat, R. Casella, T. M. Wheeler, K. M. Slawin and
S. P. Lerner, J Urol, 2003, 170, 2248–2252.
106 T. R. Griffiths, I. Brotherick, R. I. Bishop, M. D. White, D. M. McKenna,
C. H. Horne, B. K. Shenton, D. E. Neal and J. K. Mellon, Br J Cancer, 1996,
74, 579–584.
107 A. S. Protheroe, R. E. Banks, M. Mzimba, W. H. Porter, J. Southgate, P. N.
Singh, M. Bosomworth, P. Harnden, P. H. Smith, P. Whelan and P. J. Selby,
Br J Cancer, 1999, 80, 273–278.
108 R. T. Bryan and C. Tselepis, J Urol, 2010, 184, 423–431.
109 M. J. Wheelock, Y. Shintani, M. Maeda, Y. Fukumoto and K. R. Johnson,
J Cell Sci, 2008, 121, 727–735.
110 M. Przybylo, D. Hoja-Lukowicz, A. Litynska and P. Laidler, Cancer Cell
International, 2002, 2, 6.
111 M. Granovsky, J. Fata, J. Pawling, W. J. Muller, R. Khokha and J. W.
Dennis, Nat Med, 2000, 6, 306–312.
112 H. Ishimura, T. Takahashi, H. Nakagawa, S. Nishimura, Y. Arai, Y.
Horikawa, T. Habuchi, E. Miyoshi, A. Kyan, S. Hagisawa and C. Ohyama,
Clin Cancer Res, 2006, 12, 2506–2511.
113 D. Cox, M. Brennan and N. Moran, Nat Rev Drug Discov, 2010, 9, 804–820.
114 C. M. Longhurst and L. K. Jennings, Cell Mol Life Sci, 1998, 54, 514–526.
115 J. W. Tamkun, D. W. DeSimone, D. Fonda, R. S. Patel, C. Buck, A. F.
Horwitz and R. O. Hynes, Cell, 1986, 46, 271–282.
116 R. Pytela, M. D. Pierschbacher and E. Ruoslahti, Proc Natl Acad Sci U S A,
1985, 82, 5766–5770.
Carbohydr. Chem., 2012, 38, 156–175 | 173

117 D. J. White, S. Puranen, M. S. Johnson and J. Heino, Int J Biochem Cell Biol,
2004, 36, 1405–1410.
118 S. K. Akiyama, Hum Cell, 1996, 9, 181–186.
119 E. Pochec, A. Litynska, M. Bubka, A. Amoresano and A. Casbarra, European
Journal of Cell Biology, 2006, 85, 47–57.
120 M. A. Hollingsworth and B. J. Swanson, Nat Rev Cancer, 2004, 4, 45–60.
121 H. S. Silverman, S. Parry, M. Sutton-Smith, M. D. Burdick, K. McDermott,
C. J. Reid, S. K. Batra, H. R. Morris, M. A. Hollingsworth, A. Dell and A.
Harris, Glycobiology, 2001, 11, 459–471.
122 D. P. Scholfield, M. S. Simms and M. C. Bishop, BJU Int, 2003, 91, 560–566.
123 O. D. Hughes, A. C. Perkins, M. Frier, M. L. Wastie, G. Denton, M. R. Price,
H. Denley and M. C. Bishop, BJU Int, 2001, 87, 39–46.
124 O. D. Hughes, M. C. Bishop, A. C. Perkins, M. L. Wastie, G. Denton, M. R.
Price, M. Frier, H. Denley, R. Rutherford and P. A. Schubiger, J Clin Oncol,
2000, 18, 363–370.
125 T. Okegawa, M. Kinjo, S. Horie, K. Nutahara and E. Higashihara, Urology,
2003, 62, 182–186.
126 H. Kresse and E. Schonherr, J Cell Physiol, 2001, 189, 266–274.
127 E. Schonherr and R. Heinz-JüHausser, Clinical and Developmental Immunol-
ogy, 2000, 7, 89–101.
128 V. B. Lokeshwar, G. L. Schroeder, M. G. Selzer, S. H. Hautmann, J. T. Posey,
R. C. Duncan, R. Watson, L. Rose, S. Markowitz and M. S. Soloway, Cancer,
2002, 95, 61–72.
129 G. L. Schroeder, M. F. Lorenzo-Gomez, S. H. Hautmann, M. G. Friedrich, S.
Ekici, H. Huland and V. Lokeshwar, J Urol, 2004, 172, 1123–1126.
130 A. Raedler and S. Schreiber, Crit Rev Clin Lab Sci, 1988, 26, 153–193.
131 A. M. Wu, E. Lisowska, M. Duk and Z. Yang, Glycoconj J, 2009, 26, 899–913.
132 W. I. Weis and K. Drickamer, Annu Rev Biochem, 1996, 65, 441–473.
133 G. K. Ward, S. S. Stewart, G. B. Price and W. J. Mackillop, Histochem J,
1987, 19, 337–344.
134 L. Neutsch, V. E. Plattner, S. Polster-Wildhofen, A. Zidar, A. Chott, G.
Borchard, O. Zechner, F. Gabor and M. Wirth, J Urol, 2011, 186, 1481–1488.
135 H. Lahm, A. Hoeflich, S. Andre, B. Sordat, H. Kaltner, E. Wolf and H. J.
Gabius, Int J Oncol, 2000, 17, 519–524.
136 F. T. Liu and G. A. Rabinovich, Nature reviews.Cancer, 2005, 5, 29–41.
137 I. Camby, M. Le Mercier, F. Lefranc and R. Kiss, Glycobiology, 2006, 16,
137R–157R.
138 O. B. Garner and L. G. Baum, Biochem Soc Trans, 2008, 36, 1472–1477.
139 R. D. Cummings and F. T. Liu, in Essentials of Glycobiology, eds. A. Varki,
R. D. Cummings, J. D. Esko, H. H. Freeze, P. Stanley, C. R. Bertozzi, G. W.
Hart and M. E. Etzler, The Consortium of Glycobiology Editors, La Jolla,
California, Cold Spring Harbor NYEditon edn., 2009.
140 N. L. Perillo, M. E. Marcus and L. G. Baum, J Mol Med (Berl), 1998, 76,
402–412.
141 A. Danguy, I. Camby and R. Kiss, Biochim Biophys Acta, 2002, 1572, 285–293.
142 R. C. Hughes, Biochimie, 2001, 83, 667–676.
143 D. K. Hsu, R. Y. Yang and F. T. Liu, Methods Enzymol, 2006, 417, 256–273.
144 S. Nakahara and A. Raz, Methods Enzymol, 2006, 417, 273–289.
145 L. Cindolo, G. Benvenuto, P. Salvatore, R. Pero, G. Salvatore, V. Mirone,
D. Prezioso, V. Altieri, C. B. Bruni and L. Chiariotti, Int J Cancer, 1999, 84,
39–43.
146 S. H. Barondes, D. N. Cooper, M. A. Gitt and H. Leffler, J Biol Chem, 1994,
269, 20807–20810.
174 | Carbohydr. Chem., 2012, 38, 156–175

147 H. Leffler, S. Carlsson, M. Hedlund, Y. Qian and F. Poirier, Glycoconj J, 2004,
19, 433–440.
148 C. Fred Brewer, Biochim Biophys Acta, 2002, 1572, 255–262.
149 G. Canesin, P. Gonzalez-Peramato, J. Palou, M. Urrutia, C. Cordon-Cardo
and M. Sanchez-Carbayo, Tumour Biol, 2010, 31, 277–285.
150 N. Oka, S. Nakahara, Y. Takenaka, T. Fukumori, V. Hogan, H. O.
Kanayama, T. Yanagawa and A. Raz, Cancer Res, 2005, 65, 7546–7553.
151 L. Ingrassia, I. Camby, F. Lefranc, V. Mathieu, P. Nshimyumukiza, F. Darro
and R. Kiss, Curr Med Chem, 2006, 13, 3513–3527.
152 Y. Matsui, S. Ueda, J. Watanabe, I. Kuwabara, O. Ogawa and H. Nishiyama,
Cancer Res, 2007, 67, 1212–1220.
Carbohydr. Chem., 2012, 38, 156–175 | 175

Levansucrases of Pseudomonas bacteria:
novel approaches for protein expression,
assay of enzymes, fructooligosaccharides
and heterooligofructans
Tiina Alamäe,*a Triinu Visnapuu,a Karin Mardo,a Andres Mäea
and Alina D. Zamfirb,c
DOI: 10.1039/9781849734769-00176
Levansucrases are bacterial extracellular enzymes that act on sucrose pro-

ducing b-2,6-linked fructans of varied chain length. We summarize here our
up-to-date results on (i) novel methods for the production of heterologous
levansucrase proteins in Escherichia coli, (ii) biochemical characterization of
the levansucrases encoded in the genome of a tomato and Arabidopsis
pathogen Pseudomonas syringae pv. tomato DC3000, (iii) innovative chip-
based mass spectrometric analysis of oligosaccharidic reaction products,
(iv) isolation and characterization of the first mutants and structure-function
analysis of a levansucrase of P. syringae pv. tomato. Data on the levansucrase
LscA of P. chlororaphis subsp. aurantiaca will be presented for comparison.
As most important features of the levansucrases from Pseudomonas bacteria
we emphasize their ability to synthesize potentially prebiotic fructooligo-
saccharides from sucrose or raffinose and to transfructosylate a variety of
alternative acceptor sugars with production of heterooligofructans.
1 Levansucrase genes and proteins

Levansucrases are bacterial extracellular enzymes belonging to family 68 of
glycoside hydrolases according to Carbohydrate-Active Enzymes database
(http://www.cazy.org/).1 They use sucrose (a-Glc-1,2-b-Fru; GF) as a sub-
strate to synthesize b-2,6-linked fructans: short-chain fructooligosaccharides
(FOS) and also polymeric levan.2 Levansucrases catalyse following reactions:
(i) hydrolysis of the substrate (sucrose): GF þ H2O - G þ F;
(ii) polymerization: nGF þ acceptor - nG þ Fn-acceptor,
where G corresponds to glucose and F to fructose.
Levansucrases are present in Gram-positive bacteria such as Bacillus
subtilis, B. megaterium, Lactobacillus sanfranciscensis and L. reuteri, but
also in Gram-negative bacteria e.g. Zymomonas mobilis, Gluconacetobacter
diazotrophicus and some Pseudomonas bacteria. Levansucrase genes have
been found in the genomes of Pseudomonas chlororaphis subsp. aurantiaca
(also P. aurantiaca, Pca), P. fluorescens SBW25 and many Pseudomonas
syringae strains (http://www.cazy.org/GH68_bacteria.html).
a
Institute of Molecular and Cell Biology, University of Tartu, Riia 23, 51010 Tartu, Estonia.
E-mail: talamae@ebc.ee
b
Mass Spectrometry Laboratory, National Institute for Research and Development in
Electrochemistry and Condensed Matter, Plautius Andronescu 1, 300224 Timisoara, Romania.
c
Department of Chemical and Biological Sciences, Aurel Vlaicu University of Arad,
310130 Arad, Romania.
176 | Carbohydr. Chem., 2012, 38, 176–191

During recent years, we have been focusing on characterization of
levansucrases of P. syringae pv. tomato DC3000 (Pst) and their biosynthesis
products. P. syringae is a plant pathogenic bacterium, strains of which are
assigned to over 50 pathovars according to the host plant. P. syringae
infects strain-specifically bean, tomato, olive, cucumber and some other
economically important plants. Pst causes bacterial speck on tomato. This
strain has specific position among the P. syringae strains because it also
infects a popular model plant Arabidopsis thaliana with its genome sequence
available. Therefore, functional genomics can be applied to contribute to
the knowledge of plant-bacterium relationship. Data on completed and
ongoing sequencing projects of P. syringae strains can be found at http://
www.pseudomonas-syringae.org/.
Analysis of completed genomes of P. syringae pv. tomato DC3000,
P. syringae pv. syringae B728a and P. syringae pv. phaseolicola 1448A
shows that most P. syringae pathovars possess three levansucrase alleles,
two chromosomal and one plasmid-born. In case of Pst, lsc1 and lsc2 are
chromosomal and lsc3 resides on a plasmid. Strain B728a has no plasmids
and its chromosome encodes two levansucrase proteins. The length of levan-
sucrase proteins deduced from respective open reading frames of P. syringae
pathovars is either 415 or 431 amino acids (aa). Two types of levansucrases
of P. syringae differ mostly in their N-termini: the 431 aa proteins have
N-terminal extensions of 16 aa that are lacking in 415 aa proteins, otherwise
they share a high (89–99%) level of identity.3 At least one of the three lsc
genes of P. syringae DC3000 must be expressed in native host, because if
grown on sucrose-containing medium, it synthesizes levan and has a mucoid
phenotype. LscA of Pca is 424 aa long and its identity with levansucrase
proteins of Pst tomato is around 72%.
2 Heterologous expression of levansucrases from Pseudomonas syringae

pv. tomato with two different expression systems
2.1 Expression from PMAL1 using the pHIPMalprom vector
We have shown that PMAL1 – the promoter of the maltase gene (HpMAL1,
AL432586) from a methylotrophic yeast Hansenula polymorpha (synonym
Pichia angusta) functions constitutively in Escherichia coli. It is attributed to the
presence of two bacterial sigma 70-like boxes, TTGACA-N17-TAAATT and
GGTACA-N17-TATTAT in PMAL1 with high identity to sigma 70 consensus
TTGACA-N17-TATAAT (shown underlined).3,4 This feature of PMAL1 was
applied to express levansucrases of Pst in E. coli using a plasmid-born
expression system described in Visnapuu et al. (2008).3 Lsc genes were cloned
to pHIPMalprom with their native Shine-Dalgarno sequences.
Laboratory strains of E. coli do not metabolize sucrose. Thus, levansu-
crase-synthesizing E. coli will acquire a sucrose-positive phenotype – it will
grow on sucrose as sole carbon source in liquid culture and produce slimy
levan on solid media containing sucrose (see Fig. 1).3
All three levansucrase genes (lsc1, lsc2 and lsc3) of Pst expressed well in
E. coli. The level of expression depended on the host strain and on the gene
to be expressed i.e. the highest levansucrase activity (70 U/mg) was recorded
for lsc3 in E. coli HB101 (lac þ) and the lowest (7 U/mg protein) for lsc2 in
Carbohydr. Chem., 2012, 38, 176–191 | 177

Ref Lsc1 Lsc2 Lsc3
A
Fig. 1 E. coli RA11r transformants harboring pHIPMalprom-lsc1 (Lsc1), pHIPMalprom-lsc2

(Lsc2) or pHIPMalprom-lsc3 (Lsc3) have a mucoid phenotype due to levan synthesis when
grown on LB-sucrose (A) and LB-raffinose (B) agar plates, whereas on LB-glucose agar plates
they do not produce slimy polyfructan (C). E. coli transformed with empty vector pHIPMal-
prom is analyzed as a reference (Ref).
E. coli RA11r. Lsc1, Lsc2 and Lsc3 were prominent proteins in recombinant
E. coli cell extracts, whereas Lsc3 had the highest expression level com-
prising about 20% of the total soluble protein. Using the PMAL1-driven
expression system, inclusion body formation was not detected.3 Thus PMAL1
has appropriate strength to produce an adequate amount of soluble cata-
lytically active levansucrase protein in E. coli. Applying this expression
system, Lsc2 and Lsc3 proteins were purified from E. coli HB101 (lac þ)
extracts using size-exclusion chromatography on Sephacryl S-300 column
for further biochemical characterization.3,5
2.2 Expression from PT7 using the pURI3 vector

The pURI3 vector6 was used to produce wild-type and mutated Lsc3 proteins
of Pst in E. coli BL21(DE3) from the plasmid-based T7 promoter. The pURI
family of cloning vectors enables production of recombinant His-tagged fusion
proteins in E. coli without enzymatic restriction and ligation steps during the
cloning.7 Implementing the pURI3 vector system, levansucrases were produced
with N-terminal His6-tag that enabled their easy purification by Ni2 þ-affinity
chromatography.8 Extracts of E. coli BL21(DE3) expressing His-tagged Lsc3
had total levansucrase activity up to 200 U/mg that is much higher than
respective activity achieved when using the pHIPMalprom vector (see above).
3 Biochemical properties of the levansucrases: substrate specificity and

kinetic parameters
Levansucrases (EC 2.4.1.10; sucrose:2,6-b-D-fructan 6-b-D-fructosyl-
transferases) carry out sucrose hydrolysis using water for fructosyl acceptor
and transfructosylation of sucrose, FOS or some alternative acceptor
molecule. In addition to sucrose, many levansucrases react with a tri-
saccharide raffinose (a-Gal-1,6-a-Glc-1,2-b-Fru; GalGF) that is also a
178 | Carbohydr. Chem., 2012, 38, 176–191

widespread sugar in plants.9–11 Levansucrase protein of P. syringae pv.
phaseolicola was reported incapable of raffinose hydrolysis.12 Nevertheless,
our experimental data indicated that all three levansucrases of Pst use
raffinose. It was first suspected from colony phenotype of levansucrase-
expressing E. coli RA11r growing on LB-raffinose plates (Fig. 1). Regular
laboratory strains of E. coli possess melibiase (a-galactosidase) splitting
raffinose to galactose and sucrose whereas the RA11r strain is melibiase-
negative (see references in3) and has therefore suitable background to
evaluate levansucrase-mediated raffinose utilization. The RA11r colonies
expressing either lsc1, lsc2 or lsc3 of Pst produce mucoid colonies on LB
agar plates with sucrose or raffinose (Fig. 1). Levansucrase assay of
respective bacterial extracts proved release of reducing sugars from both,
sucrose and raffinose. Additionally, native polyacrylamide gel electro-
phoresis of cell extracts indicated levan production from both substrates.3
Since substrate specificity data of the LscA levansucrase of Pca were
absent in the literature, we addressed this issue. LscA, purchased from
Sigma-Aldrich, produced reducing sugars not only from sucrose but also
from raffinose and a tetrasaccharide stachyose (a-Gal-1,6-a-Gal-1,6-a-Glc-
1,2-b-Fru; Gal2GF).8 We have shown that all three levansucrases of Pst are
capable of stachyose splitting (8 and our unpublished data).
All four Pseudomonas-related levansucrases addressed here exhibit some
levan-hydrolyzing activity which is less than 1% of respective sucrose-
hydrolyzing activity with fructose detected as sole end-product (8 and our
unpublished data). Thus, levan can be considered as extracellular reserve
polymer for respective bacteria contributing to their survival in the envir-
onment i.e. on plant surface.
Km and kcat values for sucrose splitting and Ki for inhibition of sucrose
splitting reaction by raffinose of three levansucrases (Lsc2, Lsc3 and LscA)
from pseudomonads are presented in Table 1.
4 Predicted catalytic triad residues and 3D structures of Lsc1, Lsc2, Lsc3

and LscA proteins
Since now, crystallization studies of levansucrases have been focusing
mostly on respective proteins of Gram-positive bacteria – B. subtilis and
Table 1 Kinetic characteristics of sucrose and raffinose splitting by Lsc2 and Lsc3 from
P. syringae pv. tomato DC3000 and LscA from P. chlororaphis subsp. aurantiaca.
Km for sucrose kcat for sucrose kcat/Km for sucrose Ki for raffinose
Protein (mM) (min1) (M1 min1) (mM)
Lsc2 25.3 1.6 1.64 104 6.47 105 53.2 4.3

Lsc3 20.6 2.1a 1.37 104a,b 6.63 105b 47.6 5.0a
Lsc3* 18.5 2.5b 3.03 104b 1.64 106b 39.9 6.1
LscA 24.1 1.0b 4.32 102b 1.79 104b 80.8 11.8b
*His-tagged.
Molecular weights of the proteins used to calculate kcat values: 45.9 kDa (Lsc2), 47.8 kDa
(Lsc3), 49.6 kDa (His-tagged Lsc3), 47.0 kDa (LscA).
a
Data from3.
b
Data from8.
Carbohydr. Chem., 2012, 38, 176–191 | 179

B. megaterium. Only levansucrases of these bacteria have been crystallized
in complex with the substrate, sucrose or raffinose.2,13–15 Crystal structure
of LsdA of G. diazotrophicus (Protein Data Bank (PDB) code: 1W18) is the
sole one to represent levansucrases of Gram-negative bacteria.16 It should
be noted that levansucrases of Gram-positive and -negative bacteria are
rather diverse according to the sequence. For example, the sequence identity
of Lsc3 of Pst with LsdA of G. diazotrophicus is 39% whereas that with
SacB of B. subtilis is merely 19%.8 Similarly, overall sequence identity
between LsdA and SacB proteins is low – 26%.16 Despite of that, LsdA and
SacB have similar structure of five-bladed b-propeller with central acidic
pocket harbouring catalytic triad amino acids, two aspartates (Asp) and a
glutamate (Glu).13,16 Putative active site residues of Pseudomonas-derived
levansucrases predicted as in 8 are presented in Table 2.
In silico 3D structure modelling disclosed five-bladed b-propeller fold for
Lsc3 of Pst and LscA of Pca.8 The predicted topologies of Lsc3 and LscA
were highly similar except for some external loops. One of these loops
corresponds to region His90-Asp108 (Lsc3) or His81-Asp99 (LscA) that
occurs as insertion between the conserved neighbouring regions in these two
proteins (our unpublished data).
Here, we used similar ModWeb server-based modelling17 to predict 3D
structure for Lsc2 of Pst. Again, LsdA of G. diazotrophicus (PDB: 1W18)
was used as a template. Figure 2 shows predicted five-bladed b-propeller
structure of the Lsc2 with putative catalytic triad and His305, an equivalent
of a polymerization determinant His321 of Lsc38 highlighted. As Lsc1 and
Lsc3 proteins have the same length (431 aa) and almost identical sequence,
their structures are most probably highly identical.
5 Lsc2, Lsc3 and LscA produce not only levan, but also FOS that can be
detected by a novel mass spectrometric method
5.1 Characterization of the product spectrum of Lsc2, Lsc3 and LscA
The pattern of polymerization products, fructans, depends on reaction
conditions, but also on a particular levansucrase. Up to now, it is not clear
why some levansucrases, for example those of Bacillus bacteria, synthesize
mainly high-molecular levan whereas some other proteins like LsdA from
G. diazotrophicus, produce mostly FOS (see references in 8). We hypothesize
that polymerization pattern of a bacterial levansucrase may be related to
natural environment of the bacterium. For instance, LsdA of a sugar cane
endosymbiont G. diazotrophicus is solely responsible for nutrition of the
Table 2 Predicted catalytic triad amino acids of levansucrases from P. syringae pv. tomato
DC3000 and P. chlororaphis subsp. aurantiaca.
P. syringae pv. tomato DC3000

LscA of P. chlororaphis
Catalytic amino acid Lsc3a Lsc1 Lsc2 subsp. aurantiaca a
Nucleophile Asp62 Asp62 Asp46 Asp53

Transition state stabilizer Asp219 Asp219 Asp203 Asp210
Acid-base catalyst Glu303 Glu303 Glu287 Glu294
a
Data from8.
180 | Carbohydr. Chem., 2012, 38, 176–191

Fig. 2 Predicted 3D structure of Lsc2 from P. syringae pv. tomato with catalytic centre
zoomed in and catalytically important amino acids indicated. Corresponding amino acid
positions of Lsc3 are shown in brackets.
bacterium on sucrose in plant sap. LsdA produces a large amount of short-

chain FOS and only a very low amount of high-molecular levan (16 and
references therein). Accumulation of levan in sugar cane vascular tissues is
probably disadvantageous for both partners.
E. coli expressing lsc genes of Pst synthesizes slimy fructan polymer on
sucrose-containing agar plates (see Fig. 1). LscA of P. aurantiaca S-4380
forms polymer from sucrose that is a typical levan – it has high molecular
weight (7 105 Da), b-2,6 linkages between fructose residues in the main
chain and some branching through b-2,1 linkages.18 Enzymatic digestion
assay of Lsc3-produced levan predicted that levans of LscA and Lsc3 are
highly similar from this aspect.8 Analysis of the polymer produced by Lsc3
protein from sucrose revealed a 99:1 ratio of fructose to glucose that indi-
cates its high degree of polymerization (DP) and molecular weight.3
Importantly, when we first analysed transfructosylation products of Lsc2
and Lsc3 by thin layer chromatography (TLC), we did not detect FOS,
because considerably short reaction times (up to 30 min) and relatively low
substrate concentrations (up to 100 mM) were applied. Still, levan forma-
tion from sucrose and raffinose was detected in reactions conducted under
these conditions.3 If we extended reaction time up to 20 h and substrate
concentration till 1200 mM, FOS of different migration distance were
clearly detectable among reaction products of Lsc3 and LscA with sucrose
or raffinose.5,8 Figure 3 illustrates FOS production from sucrose by Lsc2,
Lsc3 and LscA proteins. LscA produces more high-DP FOS than Lsc3,
whereas Lsc2 and LscA produce somewhat less levan than Lsc3.
Sugar beet molasses as a cheap sucrose-rich substrate were also shown
suitable for FOS and levan production by Lsc3 and LscA. TLC analysis
revealed that the spectra of products formed either from molasses or sucrose
were highly similar.8
Polymerizing activity of levansucrases depends on reaction conditions
e.g. pH, temperature and substrate concentration. Accordingly, poly-
merizing activities of Lsc3 and LscA were promoted at high substrate
concentrations. At 1200 mM sucrose and 37 1C, transfructosylating
Carbohydr. Chem., 2012, 38, 176–191 | 181

Fig. 3 TLC analysis of reaction products synthesized by levansucrases from 1200 mM sucrose.
Reactions and TLC analysis were performed as in Visnapuu et al. (2009; 2011).5,8 Reference
sugars R1: fructose – F; 1-kestose – K; levan – L; R2: sucrose – S; nystose – N; levan – L.
* His-tagged protein.
activities (TA) of Lsc3 and LscA were similarly high – 76% and 73% of
fructose residues originating from reacted sucrose molecules were poly-
merized.8 At 300 mM sucrose, Lsc3 had much higher TA (43%) than LscA
(15%). TLC analysis of reaction samples indicated that at 300 mM sucrose,
LscA produces some kestose (DP 3), nystose (DP 4) and a small amount of
levan whereas Lsc3 produces more levan and FOS. Patterns of FOS syn-
thesized from 1200 mM sucrose by LscA and Lsc3 were similar, but LscA
was definitely better producer of high-DP FOS (see also Fig. 3). Compared
to LscA, Lsc3 is at least twice better levan producer at both, low (300 mM)
and high (1200 mM) sucrose concentration.8 These two proteins are there-
fore clearly different from LsdA of G. diazotrophicus which produces
1-kestose (DP 3) as a major transfructosylation product.19
Study of transfructosylating activity at pH values 5.0–7.0 revealed no pH-
dependence of this function for Lsc3 and LscA whereas high temperature
favoured the hydrolysis reaction.8
5.2 Chip-based electrospray mass spectrometry as a valuable tool for

glycobiology
The potential of the modern chip-based electrospray ionization (ESI) sys-
tems expanded significantly the area of mass spectrometry (MS) applic-
ability in glycobiology. The option for miniaturized, integrated devices for
sample infusion into MS was driven by the considerable increase in sensi-
tivity, ionization efficiency, reproducibility and throughput as compared to
classical capillary-based electrospray.20,21 However, because of the lower
ionization efficiency as compared to peptides and proteins, glycans required
reconsideration of the conditions to promote chip ionization. Moreover,
182 | Carbohydr. Chem., 2012, 38, 176–191

these conditions needed to be optimized for each class of carbohydrates.
The optimization procedures were shown to be complex and dependent on
the i) carbohydrate or non-carbohydrate type of labile attachments on the
sugar chain such as fucosylation, sialylation, sulfation, phosphorylation,
acetylation etc., ii) ionizability of the functional groups, iii) hydrophilic and/
or hydrophobic nature of molecules, iv) branching of the sugar chains and
v) the type of the aglycon in the case of conjugated glycans. Development of
effective protocols for system operation in the negative ion mode to detect
carbohydrates that form rather anions was another challenging task due to
the low ionization efficiency that ultimately leads to decreased sensitivity.
Up to now, two types of chip-based devices for ESI MS were successfully
implemented in glycobiology. The first category is represented by the out-
of-plane devices, where nozzle-like nanospray emitters are integrated onto a
silicon substrate from which electrospray is established perpendicular to the
substrate.21 Incorporated into a robot for automatic sample delivery, the
silicon chips yielded the NanoMate platform, uniquely suited to high-
throughput glycan analyses by ESI MS at remarkable flow rates (50–150 nl/
min) and sensitivities (low picomole to femtomole range). The second chip-
ESI system introduced in glycomics consists of a planar polymer microchip
embedding a microchannel at the end of which ESI is generated in-plane.22
These microsprayers provide a high stability of the spray in time, higher
tolerance to salts, improved signal-to-noise ratios at various flow rates (200–
400 nl/min) and flexibility to different ion source configurations, with the
additional advantage of cheaper production costs compared to silicon
technologies.
As a result of the very efficient ionization properties, both chips for ESI
preferentially form multiply charged ions and minimize the in-source
fragmentation of the groups attached to the sugar core through linkages
that cleave readily. Due to its versatility and unsurpassed sensitivity
NanoMate system is so far dominant chip-ESI system for glycoscreening
and sequencing. For glycomic studies, NanoMate has been coupled with a
variety of mass spectrometers21,23–31 such as quadrupole-time-of flight
(QTOF), high capacity ion trap (HCT), Fourier-transform ion cyclotron
(FTICR), Orbitrap MS and several fragmentation techniques21,24–27,29,30
like collision-induced dissociation (CID), electron-transfer dissociation
(ETD), sustained-off resonance (SORI)-CID and multistage CID MS
(MSn). NanoMate has been applied to almost all glycan and glycoconjugate
categories5,8,32–36 originating from different biological or synthetic matrices
and, under properly optimized conditions, provided in all cases highly
reproducible and sensitive analyses, detection of minor components in
complex mixtures and discovery of biomarker species, often without the
need of separation prior to MS analysis.
5.3 Identification of FOS synthesized by the levansucrases by chip-based

nanoESI HCT MS
Before our studies, some reports on conventional ESI MS application and
optimization for the analysis of levansucrase reaction products were avail-
able. It has been used to confirm methyl-fructoside formation from sucrose
and methanol by the Rahnella aquatilis levansucrase37 and to study
Carbohydr. Chem., 2012, 38, 176–191 | 183

composition of oligosaccharidic reaction products synthesized from sucrose
and its analogs by wild-type and mutant levansucrases of B. subtilis.38
Although presence of FOS among reaction products of Lsc3 and LscA
was detected by us using a traditional TLC, we intended to exert accurate
and up-to-date MS method to specify the length of the FOS species.
Experimental conditions for fully automated chip-based nanoESI, per-
formed on the NanoMate robot incorporating a 400 nozzle silicon chip for
ESI coupled to HCT MS, were optimized.5 Both, negative and positive ion
modes of the NanoMate-HCT MS platform could be used whereas formate
or phosphate and sodium additives of FOS were observed, respectively.
Those adducts of the saccharidic ions most probably originated from the
reaction buffer (McIllvaine’s buffer) and the solution for MS sample pre-
paration (it contained methanol and formic acid). Also, hydrated ions were
frequently spotted. Though ionization efficiency of neutral underivatized
oligosaccharides is limited and relative abundance of larger ions thereby
tends to be reduced, we detected FOS with DP up to 5 or 6 formed from
sucrose and raffinose by Lsc3 and LscA, respectively. To confirm the sac-
charidic origin of the isolated ions, MSn was applied by employing CID as
the fragmention technique. Depending on the sequencing conditions we
obtained product ion spectra of FOS-derived precursor ions up to MS3.5,8
The NanoMate-HCT MS method confirmed our TLC results – at high
substrate concentration FOS synthesis prevailed. Regarding the experi-
mental conditions, we obtained best-quality spectra of FOS if levansucrase
reactions were carried out at high (1200 mM) substrate concentration. It
should be noted that thermally inactivated protein, residual (unreacted)
substrate and buffer components of the reaction mixture did not hinder
FOS detection. However, a long-lasting steady spray was obtained when
reaction samples were dialysed or levansucrase reactions were conducted in
deionized water.5,8
Our study is the first report on introduction and optimization of novel
and highly sensitive chip-based nanoESI HCT MS method for the analysis
of underivatized FOS mixtures formed from sucrose and raffinose by a
levansucrase protein. Commercial prebiotic inulin-based oligofructose
mixtures (OraftisP95 and OraftisSynergy1) were analyzed to validate the
method. The spectra of commercial FOS were highly similar to those pro-
duced from sucrose by Lsc3 and LscA.5 This suggests potential prebiotic
effect of FOS synthesized by levansucrases of pseudomonads.
5.4 Chip-based nanoESI HCT MS revealed synthesis of

heterooligofructans by Lsc3 and LscA
Levansucrases and fructosyl transferases from Gram-positive bacteria can
besides sucrose transfructosylate also nonconventional acceptors and
thereby produce sucrose analogs.11,38–40 So, for levansucrases of Gram-
negative bacteria P. aurantiaca and Z. mobilis, transfructosylation of lactose
yielding bifidogenic lactosucrose has been described.41,42
We used above-mentioned rapid and straightforward chip-based nanoESI
HCT MS method 5 to screen possible fructosyl acceptors for LscA and Lsc3.
D-xylose, D-fucose, L- and D-arabinose, D-ribose, D-sorbitol, xylitol,
xylobiose, D-mannitol, D-galacturonic acid, methyl-a-D-glucopyranoside
184 | Carbohydr. Chem., 2012, 38, 176–191

Table 3 Nonconventional fructosyl acceptors of Lsc3 of P. syringae pv. tomato and LscA of
P. chlororaphis subsp. aurantiaca. The degree of polymerization (DP) of mass spectrometrically
detected products is shown. Data extracted from Supplementary Table S1 of 8.
Formula of DP of Lsc3 DP of LscA

Acceptor acceptor products products
D-arabinose (C5H10O5) 2-5 2-4

L-arabinose (C5H10O5) 2-5 2-4
D-fucose (C6H12O5) 2-5 2-4
D-sorbitol (C6H14O6) 2-5 2-4
D-xylose (C5H10O5) 2-5 2-4
D-ribose (C5H10O5) 2-5 2-4
Xylitol (C5H12O5) 2-5 2-4
D-mannitol (C6H14O6) 2-5 2-4
D-galacturonic acid (C6H10O7) 2-4 2-4
Methyl-a-D-glucopyranoside (C7H14O6) 2-5 2-4
Xylobiose (C10H18O9) 2-5 2-4
Fig. 4 (þ) chip-based nanoESI HCT MS of oligofructans in levansucrase (Lsc3) reaction

mixture containing sucrose as a substrate and D-arabinose as an acceptor molecule. Sodiated
heterooligofructans are shown on gray background and their hydrated forms are designated as
underlined. Cropped figure reproduced from Visnapuu et al., 20118 with permission from
Elsevier. Ara – arabinose; F – fructose; G – glucose; Na – sodium.
and D-glucosamine were tested for the acceptor function. While all these
substrates have molecular masses different from glucose or fructose, hetero-
oligofructans (HOF) are clearly detectable if formed in the reaction. The
samples for MS analysis were prepared and reactions were conducted as
described by us earlier.8 MS analysis of the samples indicated that all tested
nonconventional acceptors except for D-glucosamine were used by Lsc3 and
LscA. The chain lengths of HOF detected in case of every acceptor are listed
in Table 3.
Importantly, for preliminary screening of the biosynthesis products,
reaction mixtures were analyzed by MS without prior purification. Despite
of it, the mass spectra were of sufficient quality to specify HOF. We could
detect four different series of sodiated oligosaccharidic ions: 1) HOF with 1-
4 fructose residues added to the acceptor residue; 2) hydrated forms of HOF
with DP up to 5; 3) conventional FOS with DP up to 5 produced from
sucrose as a donor and acceptor and 4) hydrated species of conventional
FOS. No non-covalent complexes between the acceptor, substrate and
fructose were formed showing accuracy of the experiments. As an example,
Fig. 4 shows the mass spectra of Lsc3 transfructosylation products with
D-arabinose as an acceptor. Saccharidic origin of the detected HOF was
concluded by CID MSn.8
Carbohydr. Chem., 2012, 38, 176–191 | 185

Significantly, we showed for the first time that levansucrases can trans-
fructosylate D-sorbitol, D-galacturonic acid, D-mannitol, xylitol, methyl-a-
D-glycopyranoside and a disaccharide xylobiose.8 We presume that use of
alternative fructosyl acceptors and synthesis of HOF is probably a common
feature of levansucrases. It may be due to relaxed binding properties of the
þ 1 and further subsites of the active centre. According to the 1H and 13C
NMR spectroscopy data of previously studied HOF synthesized by levan-
sucrases of bacilli, we assume that the linkage type between a D-isomer of
nonconventional acceptor sugar and adjacent fructose residue is most
probably the same as between D-glucose and fructose in sucrose molecule
i.e. a-1,211,38 whereas the bond between following fructose residues is
most probably b-2,6 as typical for bacterial levans.38 We consider that
chip-based nanoESI HCT MS is a feasible high-throughput method to
screen possible acceptor molecules of various glycosyl transferases.
6 Isolation and novel screening methods for levansucrase mutants

6.1 Isolation strategies of Lsc3 mutants
Lsc3 protein was selected to initiate structure-function analysis of levan-
sucrases of pseudomonads using a traditional approach – isolation and
characterization of mutant proteins. First, we applied random ethylmethane
sulfonate (EMS) mutagenesis of E. coli transformants that harboured the
lsc3 gene on a plasmid pHIPMalprom-lsc3. The mutagenized transformants
were plated out to obtain a large amount of colonies. The total plasmid pool
from these colonies was used as template to amplify the lsc3 gene fragment
for recloning to pHIPMalprom. Then E. coli was transformed with this new
pool of plasmids and transformants were screened for mucoid phenotype on
agar medium with 10% sucrose (similarly as in Fig. 1). Plasmid DNA from
selected mutants was isolated and lsc3 gene was sequenced to identify the
mutations. Most of isolated mutants carried multiple amino acid-changing
mutations in the levansucrase gene. Based on position conservation ana-
lysis, some of the mutations (Asp300Asn and Thr302Pro) were chosen to
construct respective His-tagged single mutants site-directedly.8
Site-directed mutagenesis of lsc3 gene, production and purification of
respective N-terminally His-tagged mutant Lsc3 proteins was carried out
using mutation-introducing oligonucleotides and the pURI3 vector.8 We
substituted histidine (H) at position 321 with arginine (R), lysine (K), leu-
cine (L) and serine (S). His321 of Lsc3 is equivalent to His296 of Z. mobilis
LevU and Arg360 of B. subtilis SacB. These residues are crucial for trans-
fructosylation reaction by these levansucrases.2,9,43
6.2 Rapid microplate-based methods for the screening of levansucrase

mutants: using permeabilized levansucrase-expressing bacteria as catalysts
Permeabilized microbial cells can be used for the study of intracellular
enzyme activities.44 Earlier we have used microtitre plate-based semi-
quantitative enzyme assay in isolation and characterization of yeast
mutants. In this procedure, 0.1% CTAB (cetyltrimethylammonium bro-
mide) was used as a permeabilizing agent.45,46 We have shown that 0.1%
CTAB can also be used for permeabilization of E. coli cells and that it has
186 | Carbohydr. Chem., 2012, 38, 176–191

no inhibitory effect on levansucrase protein. Having this information, we
introduced a microtitre plate-based levansucrase assay on CTAB-permea-
bilized levansucrase-expressing E. coli.47
This type of assay is simple, rapid and can be robotized. All stages of this
procedure: precultivation of transformed E. coli, induction of the promoter
to switch on protein expression, cell permeabilization and evaluation of
catalytic activity can be performed on a microplate. In case of efficient
expression system, sufficient amount of protein (levansucrase) will be pre-
sent in 200 ml of bacterial culture in a microplate well to enable enzyme
activity analysis. Regarding our pURI3-based expression system, we have
used a following protocol to evaluate total (i.e. sucrose splitting) levansu-
crase activity of Lsc3 variants (our unpublished data; see the footnotew).
We have shown that permeabilized cells of levansucrase-expressing E. coli
can also be used for evaluation of FOS- and levan-producing ability of a
levansucrase.47 According to TLC analysis, reaction products synthesized
from sucrose by either cell extract of Lsc3-expressing E. coli or purified Lsc3
protein were highly similar to those produced by Lsc3-expressing permea-
bilized E. coli. In all cases, FOS with DP up to seven and also nonmigrating
products (high-DP FOS and levan) were revealed.47 Our current protocol
for this procedure is presented in the footnotez.
Levansucrases of Pst, if expressed in E. coli, are not secreted from the cell,
but remain mostly in the cytoplasm. In case of Lsc3 protein, 88% of the
levansucrase activity was located in the cytoplasm and 12% in the periplasm
of the host bacterium.3 Permeabilization by CTAB overcomes the perme-
ability barrier of the cytoplasmic membrane as well as cell envelope’s outer
membrane. Therefore we could detect a considerable amount of Lsc3 released
from intact cells after their treatment with CTAB (our unpublished data).
Some catabolism of sucrose can still take place in the periplasm of
untreated (intact) cells because sucrose will diffuse into the periplasm
w
Assay of total levansucrase activity on permeabilized cells of recombinant E. coli. Grow colonies
of levansucrase-producing E. coli overnight in LB broth on a 96-well microplate. Combine 50 ml
of microtitre plate-grown E. coli culture and 50 ml of 0.2% CTAB (cetyltrimethylammonium
bromide) in McIllvaine’s buffer (pH 6.0) in a new microtitre plate well and agitate for 10 minutes
at room temperature to permeabilize the cells. Then, add 50 ml of 0.3 M sucrose (final con-
centration 100 mM) in McIllvaine’s buffer (pH 6.0) to start the levansucrase reaction and agitate
further at 37 1C. To stop the reaction at suitable time point, e.g. in 5 min, withdraw 10 ml of the
reaction mixture, combine it in a new well with 30 ml of Tris buffer (200 mM, pH 8.3), heat the
plate at 96 1C for 5 min and cool on ice. To visualize glucose released from sucrose, add 160 ml of
Glucose Liquicolor reactive (Human GmbH, Germany) to the well and incubate the plate for
5 min at 37 1C. Measure the absorbance of red color at 500 nm (we use Tecan SunriseTM
microplate reader; Tecan Group Ltd., Switzerland, and respective software). If you analyse
mutant proteins, use the absorbance value obtained with the culture expressing the wild-type
protein for reference. The E. coli culture should be diluted in McIllvaine’s buffer prior to the
assay if catalytic acitvity is above the detection range of the microplate reader.
z
Assay of polymerization products of the levansucrases on permeabilized cells of recombinant
E. coli. Grow colonies of E. coli expressing a levansucrase protein overnight in LB broth on a
microtitre plate. Combine 40 ml of bacterial culture and 40 ml of 0.2% CTAB in McIllvaine’s
buffer (pH 6.0) in wells of another microplate and agitate 10 min at room temperature to
permeabilize the cells. Then, add 120 ml of 2 M sucrose (final concentration 1.2 M) in McIll-
vaine’s buffer (pH 6.0) to the wells, seal the plate firmly to avoid evaporation and incubate at
37 1C for 20 hours. Heat the plate for 5 min at 96 1C to stop the reaction and apply 0.5 ml of
diluted sample (we usually dilute 4 times in distilled water) to a TLC plate and proceed as
indicated in 8. Cultures of transformants should be suitably diluted in McIllvaine’s buffer if
catalytic acitvity is very high.
Carbohydr. Chem., 2012, 38, 176–191 | 187

through aspecific porins of the outer membrane and can thereby be meta-
bolized by periplasmic fraction of the levansucrase. Accordingly, intact
recombinant E. coli cells exhibited B12.5% of total levansucrase activity
measured from CTAB-permeabilized cells (our unpublished data). This
result agrees well with relative proportion of the levansucrase in the peri-
plasm. Due to periplasmic levansucrase activity, Lsc3-expressing E. coli
acquires the ability to grow on sucrose minimal medium.3
Above-mentioned methods can be used for high-throughput and sensitive
screening of large pools of mutants of various sugar-acting enzymes and can
also be adjusted for the assay of other enzyme types.
6.3 Characterization of first mutants of the Pst Lsc3 protein: His321,

Asp300 and Thr302 as residues implicated in polymerization reaction
For the first time, levansucrases of Pseudomonas bacteria have been
addressed from the structure-function aspect.8 His321 of Lsc3, an equivalent
of Arg360 of B. subtilis SacB and His296 of Z. mobilis LevU, was predicted
as polymerization determinant of Lsc3. The position of this residue is indi-
cated on a 3D model of the Lsc2 protein in Fig. 2. To verify significance of
His321, it was site-directedly mutated to Arg, Lys, Leu and Ser. This sub-
stitution had expected consequences: Km values for sucrose splitting reaction
increased up to 30 times (the affinity was reduced) and catalytic efficiency
decreased by more than 200 times. Highest negative effect on catalysis was
determined for H321S mutant. Expectedly, TA of His321 mutants was sig-
nificantly reduced – from 74% (His-tagged wild-type Lsc3) to 23% (H321L
mutant). So, His321 is certainly of catalytic importance for Lsc3 protein.8
The mutants D300N and T302P had two to three times reduced affinity
for sucrose splitting and kcat values were reduced by one third, compared to
wild-type Lsc3. TA of D300N mutant was rather similar to the wild-type
enzyme, whereas that of T302P was reduced to 47%. Reactions conducted as
in 8 were analysed for the amount of levan and total FOS. Levan production
can easily be monitored in a simple turbidity assay on microtitre plates. Our
study indicated that levan production by His321 substitution mutants was
strongly retarded. When wild-type His-tagged Lsc3 produced 7.2 mg/ml
levan, the T302P mutant synthesized less levan (6.2 mg/ml), whereas D300N
produced even more polyfructan (8.6 mg/ml) than the wild-type. Total
amounts of FOS produced by His321 substitution mutants ranged from
45.3 mg/ml (H321L) to 61.0 mg/ml (H321R). Respective value for T302P
mutant was 42.5 mg/ml and for D300N mutant 80.5 mg/ml. The wild-type
Lsc3 protein produced 107.9 mg/ml of total FOS under these conditions.
Pattern of FOS produced by the mutants revealed FOS with mobility of
kestose (DP 3) as major FOS species for H321R mutant while in case of
T302P mutant a product with mobility of nystose (DP 4) was also detected.
Intriguingly, the FOS pattern of D300N was very similar to that of LscA (see
data on LscA in Fig. 3) showing presence of longer-chain FOS species that
are not visible among polymerization products of wild-type Lsc3.8
According to sequence alignment of levansucrases, Asp300 and Thr302 of
Lsc3 belong to highly conserved region D(E/Q)(T/I/V)ER of levansucrases.
Position of Asp300 is almost invariant in levansucrases – only the Clostridium
acetobutylicum protein has Gln (Q) at that position.16 Unexpectedly, D300N
188 | Carbohydr. Chem., 2012, 38, 176–191

mutant did not differ much from the wild-type Lsc3 in our assay. Thr302 is
located next to predicted acid-base catalyst Glu303 of Lsc3 and is invariant in
levansucrases of Gram-negative bacteria. Levansucrases of Gram-positive
bacteria have mostly valine (V) or isoleucine (I) at that position.16 The
mutation of a residue equivalent to Thr302 in Lsc3 has not been described
for any of the levansucrase proteins. Our results indicate that replacement
of Thr302 with Pro significantly decreases transfructosylation, especially
synthesis of longer-chain FOS.
The active centre of levansucrases forms a funnel-shaped pocket in the
middle of the b-propeller.13,14,16 Fructose residue of sucrose molecule binds
at the bottom of the pocket and glucose residue on top of it.48 In B. subtilis
levansucrase, Glu340 and Glu342 of the DEIER motif as well as Arg360
belong to þ 1 subsite of the pocket. This subsite binds glucose residue of the
donor sucrose molecule. Arg360 and Glu340 form tight hydrogen bonds
with hydroxyls of the glucose residue fixing it in a proper orientation needed
for further reactions.2,14 We hypothesize that in Lsc3 protein, His321 and
Thr302 belong to þ 1 subsite of the substrate binding cavity. If these resi-
dues are mutated, sucrose binding is hindered that is reflected in increased
Km and reduced catalytic constant values of sucrose splitting reaction. The
þ 1 subsite is also involved in acceptor binding. Therefore His321 and
Thr302 substitution mutants of Lsc3 exhibit changed pattern of poly-
merization products: synthesis of levan and long-chain FOS was reduced
whereas short-chain FOS were still produced.
7 Concluding remarks and future perspectives

This report highlights our recent results on levansucrases of Pseudomonas
bacteria. Before we started our research, allelic spectrum of levansucrase
genes had been studied in many P. syringae pathovars and respective genes
of P. syringae pv. glycinea and phaseolicola were cloned and expressed in
E. coli.49,50 A levansucrase protein of P. syringae pv. phaseolicola was
purified and biochemically characterized by Hettwer and coworkers.12
We have been focusing on enzymology and product spectrum of levan-
sucrases encoded in the genome of P. syringae pv. tomato (Pst) with
LscA from P. chlororaphis subsp. aurantiaca (Pca) studied as a reference.
We conclude that levansucrases of Pseudomonas bacteria are from many
aspects very similar to those of other bacteria – they synthesize not only
levan, but also short-chain FOS and predictably have b-propeller fold that
is typical for levansucrases. Using a novel rapid and high-throughput chip-
based mass spectrometry method, we showed that oligosaccharidic reaction
products of levansucrases can be easily analysed without prior derivation
and even without purification. Importantly, the screening indicated that
transfructosylation of nonconventional acceptors is most probably a com-
mon attribute of levansucrases though beforehand demonstrated mostly for
the enzymes from bacilli. Our further studies will be focused on the struc-
ture-function relationships of levansucrases of Pst with first results con-
cerning catalytically relevant positions of Lsc3 protein denoted already in
this report. Having in hands a timesaving toolbox for cloning, protein
expression, enzymatic assay and reaction product analysis, we hope to get
Carbohydr. Chem., 2012, 38, 176–191 | 189

valuable information on Pst levansucrases that can be applied in bio-
technology but should also contribute to elucidation of their role for the
bacteria.
Acknowledgements
This research was funded by Estonian Science Foundation grants 7528,
9072 and SF0180088s08 to T.A., and Romanian National Authority for
Scientific Research, CNCS – UEFISCDI, project PN-II-ID-PCE-2011-3-
0047 to A.D.Z.
References
1 B. L. Cantarel, P. M. Coutinho, C. Rancurel, T. Bernard, V. Lombard and B.
Henrissat, Nucleic Acids Res., 2009, 37, D233.
2 W. Lammens, K. Le Roy, L. Schroeven, A. Van Laere, A. Rabijns and W. Van
den Ende, J. Exp. Bot., 2009, 60, 727.
3 T. Visnapuu, A. Mäe and T. Alamäe, Process Biochem., 2008, 43, 414.
4 L. Liiv, P. Pärn and T Alamäe, Gene, 2001, 265, 77.
5 T. Visnapuu, A. D. Zamfir, C. Mosoarca, M. D. Stanescu and T. Alamäe,
Rapid Commun. Mass Spectrom., 2009, 23, 1337.
6 B. de las Rivas, J. A. Curiel, J. M. Mancheño and R. Muñoz, Biotechnol. Prog.,
2007, 23, 680.
7 J. A. Curiel, B. de las Rivas, J. M. Mancheño and R. Muñoz, Protein Expr.
Purif., 2011, 76, 44.
8 T. Visnapuu, K. Mardo, C. Mosoarca, A. D. Zamfir, A. Vigants and T.
Alamäe, J. Biotechnol., 2011, 155, 338.
9 H. Yanase, M. Maeda, E. Hagiwara, H. Yagi, K. Taniguchi and K. Okamato,
J. Biochem., 2002, 132, 565.
10 S. A. F. T. van Hijum, E. Szalowska, M. J. E. C. van der Maarel and L.
Dijkhuizen, Microbiology, 2004, 150, 621.
11 J. Seibel, R. Moraru, S. Götze, K. Buchholz, S. Na’amnieh, A. Pawlowski and
H. J. Hecht, Carbohydr. Res., 2006, 41, 2335.
12 U. Hettwer, M. Gross and K. Rudolph, J. Bacteriol., 1995, 177, 2834.
13 G. Meng and K. Fütterer, Nat. Struct. Biol., 2003, 10, 935.
14 G. Meng and K. Fütterer, BMC Struct. Biol., 2008, 88, 16.
15 C. P. Strube, A. Homann, M. Gamer, D. Jahn, J. Seibel and D. W. Heinz,
J. Biol. Chem., 2011, 286, 17593.
16 C. Martı́nez-Fleites, M. Ortı́z-Lombardı́a, T. Pons, N. Tarbouriech, E. J.
Taylor, L. Hernández and G. J. Davies, Biochem. J., 2005, 390, 19.
17 U. Pieper, N. Eswar, B. M. Webb, D. Eramian, L. Kelly, D. T. Barkan, H.
Carter, P. Mankoo, R. Karchin, M. A. Marti-Renom, F. P. Davis and A. Sali,
Nucleic Acids Res., 2009, 37, D347.
18 K. H. Jang, E. K. Jang, S. H. Kim, I. H. Kim, S. A. Kang, I. Koh, Y. I. Park, Y.
J. Kim, S. D. Ha and H. K. Chul, J. Microbiol. Biotechnol., 2006, 16, 102.
19 Y. Tambara, J. V. Hormaza, C. Perez, A. Leon, J. Arrieta and L. Hernandez,
Biotechnol. Lett., 1999, 21, 117.
20 J. Peter-Katalinić, Methods Enzymol., 2005, 405, 139.
21 A. D. Zamfir, L. Bindila, N. Lion, M. Allen, H. H. Girault and J. Peter-
Katalinić, Electrophoresis, 2005, 26, 3650.
22 J. S. Rossier, N. Youhnovski, N. Lion, E. Damoc, S. Becker, F. Reymond, H.
H. Girault and M. Przybylski, Angew. Chem. Int. Ed. Engl., 2003, 3, 54.
190 | Carbohydr. Chem., 2012, 38, 176–191

23 A. D. Zamfir, S. Vakhrushev, A. Sterling, H. J. Niebel, M. Allen and J. Peter-
Katalinić, Anal. Chem., 2004, 76, 2046.
24 A. D. Zamfir, C. Flangea, D. G. Seidler, E. Sisu and J. Peter-Katalinić,
Electrophoresis, 2011, 32, 1639.
25 A. D. Zamfir, C. Flangea, A. Serb, E. Sisu, L. Zagrean, A. Rizzi and D. G.
Seidler, Methods Mol. Biol., 2012, 836, 145.
26 M. Froesch, L. Bindila, G. Baykut, M. Allen, J. Peter-Katalinić and A. D.
Zamfir, Rapid Commun. Mass Spectrom., 2004, 18, 3084.
27 R. Almeida, C. Mosoarca, M. Chirita, V. Udrescu, N. Dinca, Ž. Vukelić, M.
Allen and A. D. Zamfir, Anal. Biochem., 2008, 378, 43.
28 S. Kim, R. P. Rodgers, G. T. Blakney, C. L. Hendrickson and A. G. Marshall,
J. Am. Soc. Mass Spectrom., 2009, 20, 263.
29 C. Flangea, A. Serb, E. Sisu and A. D. Zamfir, Biochim. Biophys. Acta, 2011,
1811, 513.
30 C. Flangea, E. Sisu, D. G. Seidler and A. D. Zamfir, Anal. Biochem., 2012, 420, 155.
31 A. Ruiz-Aracama, A. Peijnenburg, J. Kleinjans, D. Jennen, J. van Delft, C.
Hellfrisch and A. Lommen, BMC Genomics, 2011, 12, 251.
32 Ž. Vukelić, S. Kalanj Bognar, M. Froesch, L. Bı̂ndilă, B. Radić, M. Allen, J.
Peter Katalinić and A. D. Zamfir, Glycobiology, 2007, 17, 504.
33 C. Schiopu, A. Serb, F. Capitan, C. Flangea, E. Sisu, Z. Vukelic, M. Przybylski
and A. D. Zamfir, Anal. Bioanal. Chem., 2009, 395, 2465.
34 A. Serb, C. Schiopu, C. Flangea, E. Sisu and A. D. Zamfir, J. Mass Spectrom.,
2009, 44, 1434.
35 E. Sisu, C. Flangea, A. Serb and A. D. Zamfir, Amino Acids, 2011, 41, 235.
36 A. D. Zamfir, C. Flangea, A. Serb, E. Sisu, N. Dinca, P. Bruckner and D. G.
Seidler, Proteomics, 2009, 9, 3435.
37 M. G. Kim, C. H. Kim, J. S. Lee, K. B. Song and S. K. Rhee., Enz. Microb.
Technol., 2000, 27, 646.
38 R. Beine, R. Moraru, M. Nimtz, S. Na’amnieh, A. Pawlowski, K. Buchholz and
J. Seibel, J. Biotechnol., 2008, 138, 33.
39 M. Tieking, M. A. Ehrmann, R. F. Vogel and M. G. Gänzle, Appl. Microbiol.
Biotechnol., 2005, 66, 655.
40 J. Seibel, R. Moraru and S. Götze, Tetrahedron, 2005, 61, 7081.
41 W. C. Han, S. H. Byun, J. C. Lee, M. H. Kim, S. A. Kang, C. H. Kim, E. Wha
Son and K. H. Jang, J. Biotechnol., 2007, 131, S113.
42 W. C. Han, S. H. Byun, M. H. Kim, E. H. Sohn, J. D. Lim, B. H. Um, C. H.
Kim, S. A. Kang and K. H. Jang, J. Microbiol. Biotechnol., 2009, 19, 1153.
43 S. Y. Li, M. Chen, G. Li, Y. L. Yan, H. Y. Yu, Y. H. Zhan, Z. X. Peng, J. Wang
and M. Lin, Acta Biochim. Pol., 2008, 55, 201.
44 T. Alamäe and A. Järviste, J. Microbiol. Methods, 1995, 22, 193.
45 T. Alamäe and L. Liiv., Folia Microbiol., 1998, 43, 443.
46 T. Kramarenko, H. Karp, A. Järviste and T. Alamäe., Folia Microbiol., 2000,
45, 521.
47 K. Mardo, T. Visnapuu and T. Alamäe, in Proceedings of the 6th International
Conference on Polysaccharides-Glycoscience, eds. R. Rapkova, J. Copikova, E.
Šarka, Praque, Czech Republic, 2010, pp. 36–39.
48 L. K. Ozimek, S. Kralj, M. J. E. C. van der Maarel and L. Dijkhuizen,
Microbiology, 2006, 152, 1187.
49 U. Hettwer, F. R. Jaeckel, J. Boch, M. Meyer, K. Rudolph and M. S. Ullrich,
Appl. Env. Microbiol., 1998, 64, 3180.
50 H. Li and M. S. Ullrich. J. Bacteriol., 2001, 183, 3282.
Carbohydr. Chem., 2012, 38, 176–191 | 191

Recent advances on the application of NMR
methods to study the conformation and
recognition properties of carbohydrates
Ana Ardá, M. Álvaro Berbı́s, Pilar Blasco, Angeles Canales,
F. Javier Cañada, Ma Carmen Fernández-Alonso,
Filipa Marcelo and Jesús Jiménez-Barbero*
DOI: 10.1039/9781849734769-00192
1 Introduction
This report gathers selected and recent examples of studies of the con-
formation, structure, dynamics and binding features of saccharides and
analogues. It has not been our intention to be exhaustive, since many groups
are working in this field either on a regular basis or making frequent con-
tributions. It has been our aim to present different examples of the applica-
tion of NMR methods and protocols for analysing saccharide conformation.
We will start by presenting new advances from the NMR methodological
viewpoint to later describe some key examples of applications of NMR for
sugar conformation determination, by discriminating between the free and
bound state and also from natural and synthetic sugars (or mimetics
thereof).
2 The access to new NMR parameters and methodological developments

The detailed analysis of chemical shifts and coupling constants combined
with the use of 13C labeled compounds give rise to important insights in the
understanding of carbohydrate conformation. In this sense, it should be
mentioned that the amide cis/trans isomerization in methyl 2-acetamido-2-
deoxy-D-glucopyranoside (MeGlcNAc) has been recently quantified by
measuring 1H and 13C chemical shifts and 1H-1H and 13C-13C spin-coupling
constants [1]. Single 13C-labeled methyl 2-deoxy-2-formamido-D-glucopyr-
anoside anomers were first characterized. Then, double 13C-labeled methyl 2-
acetamido-2-deoxy-D-glucopyranoside anomers were investigated, leading
to the detection and quantification of cis and trans amides in this key amino
sugar. These studies show that aqueous solutions of GlcNAc-containing
oligosaccharides present measurable amounts of both cis and trans amides,
which may influence their biological properties.
The accurate measure of coupling constants also allows developing
Karplus equations for conformational analysis of carbohydrates. In this
regard, eight new Karplus relationships have been recently determined for
use in conformational studies of saccharide N-acetyl side-chains in solution
by NMR spectroscopy [2]. Eight NMR vicinal spin-spin coupling constants
have been investigated, six that report on the C2-N2 bond (3JH2,NH, 3JH2,CO,
3
JC1,NH, 3JC3,NH, 3JC1,CO, 3JC3,CO) and two focusing on the amide bond
3
( JNH,CH3, 3JC2,CH3) of this side chain. Moreover, from the methodological
Centro de Investigaciones Biológicas, CSIC, Ramiro de Maeztu 9, 28040, Madrid, Spain.

E-mail: jjbarbero@cib.csic.es
192 | Carbohydr. Chem., 2012, 38, 192–214

viewpoint, it has been demonstrated that 3JHH strong coupling may strongly
affect the reliable measurement of 1JCH values. The authors concluded that
that spectral simulation yields accurate 1JCH with errors as low as 1%, even
in the presence of strong coupling [3].
On the other hand, theoretical density functional theory (DFT) methods
have been recently used to analyze the structure and the NMR spin-spin
coupling constants of a heparin disaccharide [4]. In this work, the effects of
solvation and counterions on both 1H-13C and 1H-1H coupling constants
have been investigated. The optimized disaccharide geometry showed that
the change of counterion (Ca2 þ instead of Naþ ) influenced the geometry of
pyranose rings and the glycosidic linkage conformation. The DFT-com-
puted three bond proton-proton spin-spin coupling constants agreed well
with published experimental data and indicated that the population of the
1
C4 chair form of the 2-O-sulfated iduronic acid residue increased in the
presence of Ca2 þ ions compared to the presence of Naþ ions. This analysis
also showed that the Fermi contact term was not always dominant and that
paramagnetic and diamagnetic contributions considerably influenced the
magnitudes of the proton-proton spin-spin coupling constants.
It is also worth mentioning that despite the important information
that can be obtained from coupling constants and NOE analysis, there is a
need of considering additional parameters to completely characterize
complex carbohydrates. This is specially required in those cases where the
number of long range NOEs that can be detected is very low. In these
systems, Residual Dipolar Couplings (RDCs) may provide important
structural information. In a recent work, RDCs have been used in combi-
nation with proton-proton cross-relaxation rates, trans-glycosidic coupling
constants and molecular dynamics simulations to study the complex con-
formational equilibria of a human milk pentasaccharide, LNF-1[5]. This
work also included generalized order parameters, obtained from nuclear
spin relaxation experiments. The analysis successfully showed the con-
formational space which is accessible to LNF-1.
Obviously, carbohydrates are rather flexible molecules. Therefore, NMR
observables do not always correlate with a single conformer but with an
ensemble of low free energy conformers that can be accessed by thermal
fluctuations. In this regard, a novel procedure to identify and weight the
contribution of the different conformers based on the comparison of the
experimental residual dipolar couplings with those estimated for a family of
conformers has been recently described [6]. In this work, a program based
on a genetic algorithm was developed that generates the best set of statistical
weights for the conformers derived from the conformational searches. This
algorithm was applied to derive the subensemble of conformers that appears
to be most important in the case of six different complex human milk
sugars.
RDC measurements require a partial alignment of the NMR samples,
since they average to zero under isotropic motion conditions. The first
described examples employed external alignment media, such as phospho-
lipid bicelles, lamellar phases made of cetylpiridinium chloride/hexanol/
brine mixtures or phages in aqueous solution. However, none of these
media is universally applicable since they are sensitive to pH, salt, and other
Carbohydr. Chem., 2012, 38, 192–214 | 193

experimental conditions. In addition, the media could interact with the
target molecule, depending on its chemical nature. In this sense, recent
studies have suggested to employ paramagnetic tags directly attached to the
carbohydrate for the alignment, to overcome the problems of the external
alignment media [7], following methodologies initially developed for pro-
teins. Diamagnetic proteins were converted in paramagnetic compounds
through attachment of an EDTA-based paramagnetic tag and this concept
has been extended to study carbohydrates and their complexes with receptor
molecules.
Furthermore, paramagnetic metals also cause paramagnetic shifts. These
so-called pseudocontact shifts (PCSs) arise from dipolar interactions
between the unpaired electrons of the metal ion and the nuclei in the vicinity
of this ion. They contain important structural information since they
depend on the distance between the metal ion and the nuclei. The first
conformational studies of diamagnetic carbohydrates bearing a lanthanide
binding tag have been recently described showing the potential of this new
methodology for analysing both rigid and flexible carbohydrates [8].
NMR techniques have been also developed to be able to study cell
extracts. As example, the neural cell metabolism in HT-22 cells that have
been turned into a glycolytic state, has been investigated [9]. Two different
metabolic pathways could be followed by NMR detection of UDP-GlcNAc,
the principal metabolite, in conjunction with 13C labelling of the precursor
metabolites, in this case 1-13C glucose. The application of this methodology
to other cell lines and primary cell cultures may open new avenues in this
field.
3 Applications. Saccharides in solution

3.1 Oligosaccharide and polysaccharides
Oligosaccharides display flexibility in solution, especially around the
glycosidic torsion angles and their lateral chains (responsible for the
orientations of the hydroxy groups and of the hydroxymethyl chains) [10].
Depending on the chemical natures of the constituent monosaccharides,
equilibria between different ring shapes can also be present [11, 12]. In any
case, detailed conformational studies based on molecular modelling and
NMR spectroscopy [13, 14] have allowed the deduction that some oligo-
saccharides display major conformations in solution, with local fluctuations
around the torsional degrees of freedom [15]. Computational methods, and
in particular molecular dynamics (MD) simulations, are essential com-
plementary tools to NMR data to characterise the structures and dynamics
of glycans and glycoconjugates in their free and bound states [16].
Nowadays, one relevant item in biomedical research is the synthesis of
effective antibacterial or antivirus vaccines; that is, more immunogenic and
less toxic. NMR-based conformational analysis has permitted the study of
the interaction epitope of oligo- and polysaccharides with antibodies and to
dissect the minimum structural requirement for conferring binding
affinity to these antibodies. In particular, the O-antigen (O-Ag) repeat units
(RUs) of all S. flexneri serotypes contain the tetrasaccharide backbone:
2)-a-L-Rhap-(1-2)-a-L-Rhap-(1-3)-a-L-Rhap(1-3)-b-D-GlcpNAc-(1-,
194 | Carbohydr. Chem., 2012, 38, 192–214

with glycosylation/s and/or O-acetylation defining the different serotypes.
The conformational basis of the serotype specificity has been investi-
gated for oligosaccharides representing three RUs of each serotype
[17]. The serotype-specific substitutions of the backbone did not induce
any new backbone conformation in almost all O-Ags serotype simula-
tions, although the substitutions locally restrained the accessible con-
formational space. Overall, a branched a-D-glucopyranose influenced
the conformational behaviour with flexibility restrictions of the closest
backbone glycosidic linkage on the nonreducing side for all the studied
oligosaccharides.
The conformational flexibility and dynamics of two (1-6)-linked dis-
accharides related to glycosyl transferase GnT-V has been investigated [18].
NMR NOE and T-ROE spectroscopy experiments, coupling constants
and molecular dynamics (MD) simulations were used in the analyses.
The synthesized compounds were a-D-[6-13C]-Manp-OMe derivatives,
which facilitated the determination of two- and three-bond 1H,1H, 1H,13C
and 13C,13C-coupling constants. The population distributions for the
glycosidic o torsion angle in different sugars were determined by this
method. The dynamic models, generated with the PARM22/SU01
CHARMM-based force field were in very good agreement with the
experimental observations. For instance, b-D-GlcpNAc-(1-6)-a-D-Manp-
OMe is described by an equilibrium of populated states in which F has the
exo-anomeric conformation, C shows an extended antiperiplanar geometry,
and o shows a distribution of populations between the gauche–trans and
the gauche–gauche states. The use of site-specific 13C labelling leads to
increased spectral dispersion and the combination with computational
procedures opens a promising field for structural studies of complex
carbohydrates.
An investigation of the conformational behavior of methyl b-maltoside,
methyl a-cellobioside, and methyl b-cellobioside disaccharides using NMR
spectroscopy and molecular dynamics (MD) techniques has been presented
[19]. Now, emphasis was placed on the validation of a force field for hexo-
pyranose disaccharides followed by elucidation of the conformational
properties of two different types of glycosidic linkages, a-(1-4) and b-(1-
4). Careful analysis of the data suggested that the intramolecular inter-
actions in the current force fields may be slightly underestimated, leading to
slight undersampling of the syn conformation. The proposal is that the
observed limitations come from the lack of polarization in the potential
energy. Ongoing efforts to develop next generation empirical force fields
that include electronic polarizability may allow this assumption to be
directly tested. Following this idea, the conformational space available to
a-D-Manp-(1-2)-a-D-Manp-OMe, the model for the external disaccharide
in N- or O-linked glycoproteins, has been determined using experimental
data and simulations combined with a maximum entropy approach [20].
Reparametrization of homo- and heteronuclear Karplus relationships for
glycosidic torsion angles in which the importance of electronegative sub-
stituents on the coupling pathway was deemed essential resulting in four
derived equations, two 3JCOCC and two 3JCOCH being different for the F and
C torsions, respectively.
Carbohydr. Chem., 2012, 38, 192–214 | 195

A novel method for the overexpression of 13C-labeled oligosaccharides
has been presented [21] that makes use of genetically engineered Saccharo-
myces cerevisiae cells. Furthermore, 13C labelling at selected positions of the
sugar residues can be achieved by using a site-specific 13C-enriched glucose
as metabolic precursor. This facilitates NMR spectral assignments and
provides the technical basis for NMR analysis of structures, dynamics, and
interaction of larger, branches oligosaccharides.
The synthesis of two novel carbasugar analogues of a-L-iduronic acid has
been described in which the ring-oxygen has been replaced by a methylene
group [22]. In analogy with the conformational equilibrium described for
a-L-IdopA, the conformation of the carbasugars was investigated by NMR
experiments. The carbasugar bioisosteres do not exhibit the conformational
flexibility reported for a-L-iduronic acid in oligo- or polysaccharides, which
is of great importance for its interaction with some protein receptors, such
as antithrombin.
The glycosaminoglycan (GAG) chondroitin sulfate is essential in human
health and disease but exactly how sulfation dictates its 3D-structure at the
atomic level is unclear. Homogenous oligosaccharides of unsulfated chon-
droitin (with and without 15N-enrichment) have been prepared and ana-
lyzed by NMR [23]. An experimental 3D-model of the hexasaccharide has
been presented. The unsulfated chondroitin bond geometry differs slightly
from hyaluronan by rotation about the b(1-3) dihedral angle, while the
b(1-4) linkage is unaffected. Furthermore, the comparison shows that this
glycosidic linkage geometry is similar in chondroitin-4-sulfate. It seems
that both hexosamine OH-4 and OH-6 atoms are solvent exposed in
chondroitin, explaining why this molecule is amenable to sulfation, while
hyaluronan is not. Moreover, 4-sulfation has little effect on backbone
conformation.
a-L-Rhap-(1-2)[a-L-Rhap-(1-3)]-a-L-Rhap-OMe and its site-specifi-
cally 13C-labeled(C2,C2)-isotopologue have been studied [24]. Eight con-
formationally dependent coupling constants were determined, and used
together with previously determined trans-glycosidic 3JC,H, and seven
interresidue 1H,1H-T-ROEs, for further conformational analysis of the
trisaccharide. This study highlights that different approaches are available
in order to obtain all the information needed for the molecular system under
investigation. The ABO histo-blood group antigens are best known for their
important roles in solid organ and bone marrow transplantation as well as
transfusion medicine. The synthesis of the ABO type III and IV antigens
with a oct-7-en-1-yl aglycone has been reported [25], along with the NMR
conformational study of the ABO type I to VI antigens. These NMR
investigations showed very little difference in the 1H chemical shifts, as well
as 1H–1H coupling constants, across all compounds, suggesting that these
ABO subtypes adopt nearly identical conformations in solution. One
exception is the chemical shift for H5 of the fucose residue of the A type V
tetrasaccharide, which resonated around 4.65 ppm, a deviation of 0.33 ppm
from the 4.32 ppm average. The significance of this observation is currently
unclear.
Combination of 1H, 13C and 39K NMR spectroscopy with single-crystal
X-ray crystallography have allowed to study the complex formation of
196 | Carbohydr. Chem., 2012, 38, 192–214

1,6-anhydro-b-maltose and potassium ions [26]. In solution, diffusion
coefficients and 39K-T1 data unambiguously indicated complex formation,
while 13C NMR studies indicated the importance of the 1,6-anhydro moiety
for the complex formation, also supported in the X-ray crystal structure.
The description of the re-orientational dynamics of flexible molecules is a
challenging task, in particular when the rates of internal and global motions
are comparable. The commonly used simple mode-decoupling models are
based on the assumption of statistical independence between these motions.
This assumption is not valid when the time scale separation between their
rates is small, a situation that was found to arise in oligosaccharides in the
context of certain internal motions. To make possible the interpretation of
NMR spin relaxation data from such molecules, a comprehensive approach
generally applicable to flexible rotators with one internal degree of freedom
has been developed [27]. The method has been successfully tested on b-D-
Glcp-(1-6)-a-D-[6-13C]-Manp-OMe dissolved in DMSO-d6/D2O.
Passing to polysaccharides, the capsular polysaccharide of Escherichia
coli K5 has been hypothesised to promote virulence through its molecular
mimicry of host heparan sulphate. To test this hypothesis, pure oligo-
saccharides from K5 have been produced and their conformational prop-
erties investigated with ultra-high-field NMR (900 MHz) [28]. All
carbohydrate rings adopt a 4C1 conformation, the amide side chains have a
trans orientation and the hydroxymethyl group is freely exposed to bulk
solvent. Initial models of the glycosidic linkage conformation based upon
simple interpretation of NOE cross-peaks suggests that the b(1-4) linkage
adopts a 3D geometry of FE601, CE01 and the a(1-4) linkage prefers
FE–301, CE–301. In this conformation, the overall molecular geometries
of K5 polysaccharide, heparan sulphate and even fully-sulphated heparin
are remarkably similar. These results substantiate the hypothesis that the
K5 capsular polysaccharide confers virulence to E. coli K5 by being a 3D
molecular mimetic of host heparan sulphate, helping it to evade detection
by the mammalian immune system.
The conformational features of hyaluronic acid [4)-D-b-GlcA-(1-3)-D-b-
GlcNAc-(1-] have also been investigated using NMR and molecular
modelling [29]. A decasaccharide fragment of sodium hyaluronate (HA) was
subjected to 3.5 ns of molecular dynamics in explicit water and the approach
consisted in the measurement of NMR residual dipolar coupling (RDC),
which were used to filter the molecular dynamics data, by retaining those
structures which were in agreement with the experimental observations. As
a result, this protocol led to the conclusion that hyaluronic acid can adopt
two different arrangements, which can be described by three- and four-
folded left-handed helix, with the first one as the major conformer. These
two different conformers coexist in solution and convert into each other.
This information evidences how subtle variations in the dihedral angles are
accompanied by a switch from one geometry to another (from a three- to a
four-folded helix).
The use of lactic acid bacteria in milk fermentation results in
favorable physical and rheological properties due to exopolysaccharide
(EPS) production [30]. The hexasaccharide repeating unit EPS from
S. thermophilus ST1 has been elucidated and presents the following
Carbohydr. Chem., 2012, 38, 192–214 | 197

structure: 3)[a-D-Glcp-(1-4)]-b-D-Galp-(1-4)-b-D-Glcp-(1-4)[b-D-Galf-
(1-6)]-b-D-Glcp-(1-6)-b-D-Glcp-(1-. Thus, the EPS consists of a
backbone of four sugar residues with two terminal ones making up two side-
chains of the repeating unit. The use of efficient NMR techniques based on
parallel acquisition and TILT experiments was very valuable for the
structural analysis.
NMR has been employed to study transient hydrogen bonding in a
polysaccharide dissolved in water without any cosolvent at room tem-
perature. In this case, the pentasaccharide repeating unit of the Escherichia
coli O142 lipopolysaccharide has been used as model system [31]. A 105 ns
MD predicted transient inter-residue hydrogen bonds from the GalNAc
NH groups. To experimentally investigate these predictions, the poly-
saccharide was uniformly 13C,15N-enriched and the NH, carbonyl, C2, C4,
and methyl resonances of the GalNAc and GlcNAc residues assigned.
Temperature dependence of the amide NH chemical shifts and one-bond
NH J couplings supported that NH groups on two of the GalNAc residues
are donors in transient hydrogen bonds. The remaining GalNAc and
GlcNAc NHs do not appear to be donors These results raise the biologically
important possibility that carbohydrates may utilize hydrogen bonding to
assemble three-dimensional scaffolds that would form the basis of biomo-
lecular carbohydrate recognition and function.
Strong efforts have been developed for implementing automatic struc-
tural determination of polysaccharides [32], by using the computerized
CASPER approach, based on 1H and 13C chemical shifts.
3.2 Glycopeptides
Glycosylation is one of the most widespread post-translational modifica-
tions of proteins [33]. Thus, the conformational study of glycopeptides and
peptidoglycans continues to be a hot topic of research. The most prevalent
type of O-glycosidic linkage involves an a-N-acetylgalactosamine (GalNAc)
unit and the b-hydroxyl group of either Ser or Thr at the peptide chain [34].
This glycosylation pattern is commonly present in antigenic mucins [35, 36].
In the last few years, the combination of experimental NMR data (chemical
shifts, coupling constants and NOEs) assisted by molecular modelling
protocols has strongly contributed to disclose the effect of a-O-glycosyla-
tion on the conformation of mucin-type peptides [37]. The extensive work of
Live and co-workers revealed that a-O-linkages induce the peptide back-
bone to adopt an extended conformation [38]. More recently, the work of
Corzana et al. has demonstrated that the nature and sequence of the amino
acid aglycone also modulates the conformation and dynamics effects of Tn
antigen’s clusters [39, 40]. In addition, Nishimura’s group has combined MS
and NMR techniques to study the conformational effects of O-glycosylation
on the peptide backbone structures during enzymatic mucin domain
assembly in vitro by using an isoform of UDP-GalNAc:N-acetyl galacto-
saminyl transferase-T [41]. Structural analysis, combining NOESY data
with structural refinements of the synthetic key intermediates uncovered
that the preferential installation of a-GalNAc at Thr-10 of MUC4 is a
crucial step for forcing the stability of the peptide backbone structure at this
198 | Carbohydr. Chem., 2012, 38, 192–214

area. The authors proposed that this event may provide a preferred sub-
strate conformation for the subsequent multiple glycosylation.
Additional investigations have focused on the possibility that the biolo-
gical activity of the N/OFQ peptide, a native ligand of the pain-related and
viable drug target NOP receptor could be modulated by glycosylation. The
relationship with conformational effects was also addressed [42]. Hence, the
pharmacological analysis of three novel glycopeptides [Thr5-O-a-D-Gal-
NAc-N/OFQ] (1), [Ser10-O-a-D-GalNAc]-N/OFQ (2) and [Ser10-O-b-D-
GlcNAc]-N/OFQ] (3) revealed that glycopeptide 2 exhibited an improved
binding affinity, whereas 1 showed a remarkably reduced binding affinity
compared to the parent compound and to 3. NMR data complemented with
modelling allowed identifying a certain degree of helical structure for the
parent N/OFQ peptide, as well as for 2 and 3 in micellar solutions. In
contrast, for 1, a distinct set of remote NOE contacts were detected, sug-
gesting that this glycopeptide exhibited a significant population of folded
hairpin-like structures. Thus, the authors proposed that linear helical
structures of nociceptin analogues are more complementary and interact
more efficiently with the NOP receptor than folded structures.
The stable incorporation of 13C and 15N into glycoconjugates has allowed
to employ NMR methods to characterize the dynamics and interactions of
glycoprotein glycans, as recently reviewed [43]. The linewidth of the 13C
anomeric peak is inversely proportional to the transverse relaxation rate,
T2, and permits the analysis of the mobility of the IgG-Fc glycans. In
particular, the data obtained for the Fc glycoprotein showed that two
galactose peaks exhibited extremely different line widths, indicating that
these sugar residues have significantly different mobilities in Fc glycans [44].
In addition, the solvent accessibility/hydrogen bonding interactions of the
GlcNAc side chains in Fc was monitored by amide exchange experiments.
Slow rates for amide proton exchange are associated with shielding from
solvent, usually through hydrogen bonding interactions. After selective
isotope labelling of the N-glycans on the Fc fragment of IgG, and com-
bining relaxation dispersion-type experiments with temperature-dependent
chemical shift perturbation measurements, it was demonstrated that both
[a-(1-3)- and a-(1-6)Man] glycan branches of the antibody are
accessible to receptors and rather mobile in solution [45].
Conformational studies on unusual phosphorylated O-mannosyl glyco-
peptides from a-dystroglycan have also been presented [46]. It was found
that a glycopeptide having a 6-phospho-O-mannosyl residue is not an
acceptor for POMGnT1, which attaches b(1-2)-GlcNAc to O-mannosyl
moieties. In addition, the conformational analysis by NMR showed that the
O-mannosyl modification did not exert a major conformational effect on the
peptide backbone. It was then proposed that these residues, introduced at
the early stages of glycoprotein glycosylation, have an ability to regulate the
loci of subsequent O-GlcNAc additions through conformational effects.
One additional example has focused on why PrkC Thr-kinase induces
germination of bacterial spores in response to peptidoglycans of the DAP-
type but not to those of the Lys-type cell wall. Thus, the structural
requirements necessary for recognition and binding has been addressed using
STD-NMR and TR-NOESY, together with mutagenesis experiments [47].
Carbohydr. Chem., 2012, 38, 192–214 | 199

In fact, the extracellular region of PrkC senses the peptidoglycan through
the DAP-moiety-mediated interactions with an arginine residue of the
EC-PrkC receptor.
The conformational properties of the Glc3Man unit has been studied to get
information on the chaperon-assisted glycoprotein folding pathway [48].
Calnexin recognises the terminal Glca1-3Mana linkage, formed by trim-
ming of the Glca1-2Glca1-3Glca1-3Mana (Glc3Man) unit in Glc3-
Man9GlcNAc2. Thus, the conformation of a series of N-glycans; i.e.
Glc3ManOMe, Glc3Man4,5,7GlcNAc2 and Glc1Man9GlcNAc2 using 2D
NMR NOESY, ROESY, T-ROESY and residual dipolar coupling experi-
ments in a range of solvents, along with MD simulations of Glc3ManOMe. A
single conformation for the Glca1-2Glca and Glca1-3Glca linkages was
found, along with two conformers for the Glca1-3Mana linkage. Model-
ling of the binding of Glc1Man9GlcNAc2 to calnexin suggests that the minor
conformer is recognized by calnexin. This is the first evidence presented on
glycoprotein folding that suggests that this process may be optimized to
balance chaperone-assisted and chaperone-independent pathways.
As mentioned before, stable-isotope-labelling techniques enhance the
quality of NMR data. A stable-isotope-assisted NMR analyses using
immunoglobulin G as a model system in order to gain information on the
dynamics and interactions of glycoconjugates has been reported [49]. A
series of glycopeptides cleaved from the metabolically 13C/15N-labeled IgG
glycoprotein can be used as tools for detailed NMR analyses and for
characterisation of sugar-binding modes of lectins.
3.3 Glycomimetics
The design and employment of glycomimetics as possible therapeutic agents
remains an attractive area of research [50]. NMR studies have contributed
to disclose their conformations in the free and bound state and to provide
insights on their interactions with receptors [51]. Some glycomimetics have
been employed to disrupt the formation of sugar–protein complexes,
including glycosidases and glycosyltransferases [52, 53].
As example, the molecular basis for inhibition of GH84 glycoside
hydrolases by substituted seven-membered ring iminoalditols (azepanes)
has been reported [54]. This work pointed out the importance to design
chemically stable inhibitors, but conformationally flexible. NMR data
combined with molecular mechanics (MM) and MD simulations revealed
that the polyhydroxylated azepanes adopted a distinct type of conforma-
tions in solution than those found by X-Ray crystallography at the enzyme
active site. These results showed that the enzyme selects the proper con-
formation of the conformationally flexible azepane into the enzyme active
site to optimize the inhibitor-enzyme interactions. As additional example,
the binding and substrate specificity of the recognition of b-D–mannosides
and b-D-glucosides by barley (HvBII) and rice (Os3BGlu7) b-D–glycosidases
has been investigated by NMR with quantum mechanics/molecular
mechanics (QM/MM) calculations and docking protocols [55]. STD-NMR
and TR-NOESY experiments performed using competitive thioglycoside
mimetics (4NP-S-Glc and 4NP-S-Man) revealed a distinct bioactive con-
formation for 4NP-S-Glc (B3,O and 1,4B) and 4NP-S-Man (4C1) when
200 | Carbohydr. Chem., 2012, 38, 192–214

bound to HvBII, and the same type of 4C1 conformation when bound to
Os3BGlu7. In fact, the experimental and computational data suggested that
glycoside recognition and substrate specificity of Os3BGlu7 and HvBII rely
on a combination of factors involving (i) the inherent conformational
flexibilities of the gluco- and manno-configured substrates at the active sites,
(ii) the subtle differences in the spatial disposition of the active site residues
and their capacities to form interactions with specific groups of the sub-
strates, and (iii) the small variations in the charge distributions and shapes
of the catalytic sites.
Fragment-based methods have been increasingly adopted in the drug-
discovery process, also in the carbohydrate field. The identification of the
better way of joining the different molecular fragments to constitute the
final lead continues to be a challenge [56]. Thus, techniques that permit a
dynamic-system-based drug-design for the identification, synthesis, and
evaluation of new leads offer a way to solve this problem. Ramström and
coworkers have demonstrated that 1H STD-NMR spectroscopy can be
employed for efficient and direct in situ identification of b-galactosidase
inhibitors within a virtual dynamic hemithiocetal system [57]. The best
molecular fragments identified from the dynamic approach were then
employed to derive a good b -galactosidase inhibitor using a regular syn-
thetic approach [58].
In the lectin field, different mannosyl glycomimetics have been designed
to efficiently target the extracellular lectin domain of DC-SIGN by using a
multidisciplinary approach [59, 60]. More recently, the same authors have
reported the employment of a a-fucosylamide-based mimic of Lewis X [61].
Interestingly, the subsequent insertion of an aromatic ring to replace the
Gal moiety allowed improving the binding affinity, as confirmed by STD
competition experiments [62].
An additional example includes a simple change in the chemical nature of
one glycosidic linkage to perturb the conformational properties of oligo-
saccharides and even to derive a moderate competitive inhibitor. In parti-
cular, the synthesis of the b(1-3) analogue of pentaacetyl chitopentaose,
replacing the natural b(1-4) linkage by the b (1-3) alternative has been
reported. NMR and modelling studies permitted to verify that the mimetic
is recognized by a lectin (wheat germ agglutinin) and one enzyme (a model
chitinase) [63]. The existing subtle differences in the conformational beha-
viour of both molecules allowed the mimetic still to interact with the two
proteins. However, it was not hydrolyzed by the enzyme and acted as a
moderate inhibitor of chitin hydrolysis.
One key feature in protein-carbohydrate interactions involves multi-
valency, which notably improves the affinity and specificity of biomolecular
interactions [64]. b-C-galactopyranosyl mimetics have been incorporated
within a multivalent strategies to design a potent bacterial lectin inhibitor
[65]. The multivalent C-galacto-conjugated architectures were synthetised
and evaluated against PA-IL lectin (Lec A) from pathogenic Pseudomonas
aeruginosa. NMR diffusion experiments allowed estimating the size and
topology of the multivalent compounds, indicating the expected decrease of
the diffusion coefficients with increasing diameter, as the number of per-
ipheral epitopes was increased.
Carbohydr. Chem., 2012, 38, 192–214 | 201

A new class of GlcNAc mimics for E-selectin antagonists have been
designed and synthesised [66]. The mimic consists of a cyclohexane ring
substituted with alkyl substituents adjacent to the linking position of the
fucose moiety. By STD-NMR experiments it could be shown that the
increase in affinity does not result from an additional hydrophobic contact
of the alkyl substituent with the target protein E-selectin, but rather from a
steric effect stabilizing the antagonist in its bioactive conformation. When
the GlcNAc mimetic with two methyl substituents was used the affinity was
regained. As predicted by MD simulations, equatorial alkyl groups at the
2-position in the carbocyclic GlcNAc mimetic restrict the conformational
flexibility of the core of the antagonist and entropically improve binding
affinities.
A carbohydrate peptide mimic of the O-polysaccharide of Shigella flex-
neri Y, MDWNMHAA, when bound to its complementary antibody mAb
SYA/J6, has been studied [67]. Its conformation in the bound state has been
characterised as well as the presence of immobilised water molecules in the
active site. WaterLOGSY experiments were used in conjunction with STD
methods to probe the existence of these immobilised water molecules in the
complex. MD simulations were then used to specify the likely locations of
the water molecules. Of note, those waters involved in providing com-
plementarity between the peptide and mAb SYA/J6 remained throughout
the course of the simulation.
4 Applications. The interaction of saccharides with other natural and

synthetic molecules
In the last years, a number of studies on the conformational features of the
interaction between proteins and saccharides have been reported, providing
new details, at atomic resolution, on these key processes. NMR experiments
have been employed to analyse the interaction of sugars with different
biomolecules, including proteins (lectins, antibodies and enzymes), nucleic
acids and even other saccharides. Synthetic receptors, also dubbed artificial
lectins, have been employed to model sugar interactions. Finally, a few cases
of the use of cyclodextrins as receptors will be mentioned, as examples of
sugars being receptors instead of organic molecules’ ligands.
In principle, there are different ways to characterise the complexes: On
one hand, in favourable cases, it is possible to measure the NMR para-
meters of the receptor in the free and bound states, i.e., differences in
chemical shift and NOE-contacts, before and after the formation of the
complex. On the other hand, it is usually more feasible to observe and detect
changes in the ligand NMR signals before and after the binding process. In
this latter case, the use of STD (saturation transfer difference) and TR-NOE
(transferred NOE) experiments is of paramount importance.
4.1 Carbohydrate-protein interactions

A carbohydrate-binding protein can perform conformer selection, which
has been demonstrated in different instances by combinations of NMR and
modelling [68]. Combined NMR/modelling approaches have allowed this
principle to be explored along with determinations of the specificities,
202 | Carbohydr. Chem., 2012, 38, 192–214

affinities and structural aspects of many receptor–sugar interactions both
with natural sugars and with glycomimetics. Depending on the architecture
of the protein’s binding site, the major conformation of an oligosaccharide
existing in solution might be recognised by the receptor, but in other cases, a
conformational selection process can take place with exclusive recognition
of one of the conformers [69] which might be drastically different from the
major one existing in water solution.
The current state-of-the-art NMR methodologies for the study of pro-
tein-carbohydrate interactions have been outlined in a recent review [70].
The article described the most prominent NMR techniques utilised nowa-
days, both from the ligand and receptor perspectives, as well as their
combination with molecular modelling procedures. Among the ligand-
detected NMR methods, STD and TR-NOESY experiments are regarded as
particularly useful for scoping carbohydrate-protein interactions, thanks to
the information that they provide, and to the fact that the optimum kinetic
equilibrium for such experiments, lying between the mM to low mM range,
are satisfied by the usually moderate to weak carbohydrate-protein inter-
action affinity regimes. The authors ended up by paying attention to the 3D
view of protein-carbohydrate complexes as obtained with the help of
computational protocols, such as docking and molecular dynamics, and
reported on the advent of more sophisticated computational approaches,
such as the application of QM/MM and metadynamics for exploring
mechanistic aspects of glycosidases and glycosyltransferases.
STD studies are of pre-eminent importance for investigating protein-
ligand interactions. However, accurate quantitative analyses of these
interactions have been precluded by the complex dependence of the STD
signal magnitude on a variety of factors, which introduce unacceptable
biases when dissociation constants (KD) are attempted to be estimated
through STD-titration experiments. A recent paper sought to study the
relative impact of such dependencies on the STD signal and to develop a
straight-forward protocol to directly measure the KD of protein-ligand
complexes by STD titration curve [71]. The authors identified the accu-
mulation of saturation in the free ligand through ligand rebinding events
and the longitudinal relaxation as the main sources of error when estimating
affinities through STD data. To overcome these drawbacks, they proposed a
novel method, in which the binding isotherms are constructed by using the
initial growth rates of the STD amplification factors, taking profit that, at
zero saturation time, neither rebinding events nor relaxation occurs. The
proposed protocol was validated with two well-studied carbohydrate-pro-
tein interaction model systems: the binding of GlcNAc and GlcNAcb(1-
4)GlcNAc (chitobiose) to the WGA lectin. In all instances, the affinities
estimated with this novel method were in good agreement with previously
reported data obtained with calorimetric studies.
NMR spectroscopy has also been used to characterise the binding
properties of 2G12, the only carbohydrate-directed monoclonal antibody
against the human immunodeficiency virus (HIV), which has specificity for
a cluster of mannose-rich glycans on the viral glycoprotein gp120 [72]. In
particular, the authors explored the recognition of mannose-coated gold
nanoparticles by 2G12, using SPR and STD-NMR. Since both the antibody
Carbohydr. Chem., 2012, 38, 192–214 | 203

and the nanoparticles are large systems, direct evaluation of the KD by STD
was not possible. Instead, affinity data were obtained by monitoring the
competition of the nanoparticles against previously characterized 2G12/
oligomannoside systems, revealing affinities in the low micromolar range for
the 2G12/nanoparticles recognition. Furthermore, the nanoparticles were
shown to compete with gp120 for the antibody, effectively blocking the
2G12-mediated cellular immunity to the virus in vitro. These studies are
meant to contribute to the development of mannose-gold nanoparticles as
safe, potent and versatile agents to promote immunization against the HIV
by eliciting 2G12-like antibodies.
Using a similar approach, STD studies have contributed to yield affinity
and fine specificity data for chimeric IgE antibodies towards Gala(1-3)Gal,
considered of outstanding relevance in the context of xenotransplantation
and vaccine design, as part of a study devoted to the controversial ana-
phylaxis mediated by immunogenic carbohydrate determinants [73].
Among carbohydrate-protein interactions, the specific recognition of
sugars by galectins stands among those most well-studied. Although
galectins are, by definition, lectins with specificity for b-galactosides, some
data have indicated that galectin-1 is able to interact with a-galactosides as
well. The recognition of Gala(1-3)Gal, Gala(1-4)Gal and Gala(1-6)Glc
(melibiose) by galectin-1 was studied by NMR spectroscopy [74]. STD data
revealed a pre-eminent recognition of the non-reducing (Gal) end in all
instances, while 1H-15N HSQC experiments revealed that all disaccharides
bound to the canonical recognition site, although with affinities more than
two orders of magnitude weaker than that of lactose.
Similar studies, involving epitope mapping and affinity evaluations by
1
H-15N HSQC titration experiments, have been performed on the cyano-
bacterial Oscillatoria Agardhii agglutinin (OAA) to study its binding cap-
abilities to the mannose-rich clusters found in the HIV gp120 glycoprotein
[75]. The lectin was revealed to recognise the branched core unit of Man-9
via two distinct binding sites located at opposite ends of the protein. X-ray
studies demonstrated that, unlike most HIV-inactivating lectins, OAA
recognised the core structure of the triantennary Man-9, rather than its
reducing or non-reducing ends, thus raising the question whether a syner-
gistic action with terminal mannose-recognising lectins could be explored.
Regarding the interaction between enzymes and their carbohydrate
substrates, ST6Gal-I, a glycosyltransferase of medical interest, which cat-
alyzes the transfer of sialic acid residues from CMP-NeuAc to galactose
residues of glycolipids and glycoproteins on the surface of mammalian cells
has been monitored [76]. Its structural determination by X-ray diffraction or
NMR has been precluded by the enzyme being glycosylated and membrane-
bound, and because of its large size of 40 kDa. Thus, the authors elegantly
addressed the issue of how structural information can be obtained by NMR,
even without complete structural determination. For this purpose, they
investigated several aspects of the enzyme active site from the point of view
of its bound ligand, using spin-labelled derivatives of both donor and
acceptor moieties. Previously, they had utilised CMP-4-carboxy-TEMPO, a
spin-labelled donor mimic, containing a paramagnetic nitroxide group [77].
This moiety significantly increases relaxation of nearby magnetic nuclei in a
204 | Carbohydr. Chem., 2012, 38, 192–214

r6 dependent manner and can be used to map the position of sixteen 15N-
labeled phenylalanines relative to the position of the donor. The authors
extended the study to a spin-labelled acceptor mimic, LacNAc-TEMPO,
and the results were compared to those obtained with the spin-labelled
donor. In addition, the labelled donor was used to perturb the signals of a
non-labelled acceptor, providing distance restraints between the donor
nitroxide and the acceptor carbon atoms. TR-NOE experiments were per-
formed to determine the conformation of the bound molecules, and STD
data gave information on the proximity of the ligands epitopes to the
protein surface. Taken together, these data were used to generate a model of
the enzyme active site geometry, thus providing key insights into a
mechanism of action.
The quest for suitable and selective inhibitors for glycosyltransferases is
still challenging. A novel concept that allows the design of inhibitors based
on the structure of the donor substrate binding pocket has been presented
[78]. As a first step, the design, synthesis and analysis of inhibitors of the
human blood group B galactosyl-transferase (GTB) was presented. In silico
docking of bicyclic heteroaromatic ligands to GTB and experimental ver-
ification of their binding affinities by STD gave 9-N-pentityl uric acid
derivatives as non-ionic mimics of UDP. As demonstrated, pentitol linked
uric acid derivatives are very well suited to replace uracil, the ribofuranoside
and the pyrophosphate group of UDP, resulting in a similar binding affinity
in the presence of Mg2 þ to GTB.
Following on the design and synthesis of glycosyltransferase inhibitors, a
major obstacle has always been the demanding chemistry. Glycosyl-
transferases usually possesses two distinct binding sites, one for the donor
substrate, and one for the acceptor substrate. One approach to glycosyl-
transferase inhibitors is to chemically link donor site and acceptor site
ligands to generate high affinity binders. In this context, a novel approach to
identify acceptor site ligands from a fragment library has been described
[79]. As model, galactosyltransferase (GTB) was chosen. The approach used
a combination of STD NMR, spin-lock filtered NMR experiments and
surface plasmon resonance measurements. Molecular fragments from a
fragment library that bind to the acceptor site of GTB with affinities of the
order of a natural acceptor substrate have been identified. They allow for
straightforward chemical modifications and, therefore, could serve as
scaffolds for potent GTB inhibitors.
STD-NMR studies for GlfT2 have been reported, using two trisaccharide
acceptor substrates, b-D-Galf-(1-6)-b-D-Galf-(1-5)-b-D-Galf-O(CH2)7CH3
and b-D-Galf-(1-5)-b-D-Galf-(1-6)-b-D-Galf-O(CH2)7CH3, as well as the
donor substrate for this enzyme, UDP-Galf [80]. Competition STD-NMR
titration experiments and saturation transfer double difference (STDD)
experiments with the two trisaccharides were undertaken to explore the
bifunctionality of this enzyme, in particular to answer whether one or two
active sites are responsible for the formation of both the b-(1-5)- and b-(1-
6)-Galf linkages. It was demonstrated that both bind competitively at the
same site. This fact suggests that GlfT2 has one active site pocket capable of
catalyzing both b-(1-5) and b-(1-6) galactofuranosyl transfer reactions.
The addition of UDP-Galf to GlfT2 in the presence of either one or the other
Carbohydr. Chem., 2012, 38, 192–214 | 205

trisaccharide generated a tetrasaccharide product. This observation means
that the enzyme was catalytically active under the conditions at which the
STD-NMR experiments were carried out. Moreover, epitope mapping
demonstrated a greater enhancement toward the ‘reducing’ ends of both
trisaccharides, and that UDP-galactofuranose (UDP-Galf ) made more
intimate contacts through its nucleotide moiety. This observation is con-
sistent with the greater flexibility required within the active site of the reaction
between the growing polymer acceptor and the UDP-Galf donor.
The oligomerisation profile of chemokines is also a topic of major
interest, especially in what concerns their binding to glycosaminoglycans
(GAGs). There is sparse information of such events in solution, probably
due to the poor solubility of the resulting aggregates. The structure of
CCL27, a cutaneous T cell-attracting chemokine, its oligomerisation state
and its interactions with GAGs have been investigated by NMR [81]. The
structure of CCL27 was solved by using standard procedures, and PFG
diffusion and 15N relaxation data revealed that CCL27 displays transitions
between monomer, dimer, and tetramer states within the low millimolar
range. To characterise the dimer interface, intersubunit NOEs were scoped
by utilising a 13C/15N-separated and filtered NOESY experiment on a mixed
isotope sample containing 50% 13C-labelled CCL27, and 50% 15N-labelled
CCL27, which maximised the detection of intersubunit NOEs between
1
H-13C and 1H-15N protons. The oligomeric profile of CCL27 was shown to
be more dynamic than that of other members of the family, with more than
one type of dimer being present. Direct study of GAG-promoted oligo-
merisation of CCL27 was impossible due to precipitation of the chemokine
upon addition of the heparin oligosaccharides, a fact that was attributed to
the formation of insoluble, higher order oligomers. Such effect was then
investigated on CCL2, a related chemokine. Addition of a heparin octa-
saccharide to a CCL2 sample at 1:2 ratio shifted the Ds value of the latter
from the monomer value to the expected value for a tetramer, strongly
supporting the notion of chemokine association upon interaction with
GAG.
The quaternary structure of CCL5/Rantes chemokine has also been
investigated in solution by a multidisciplinary approach involving SAXS
and NMR [82]. 1H-15N HSQC experiments on oligomeric samples revealed
only one set of cross-peaks, suggesting that the oligomeric complexes are
symmetric. In order to solve the oligomeric structure, a different strategy
was used, which involved residual dipolar coupling (RDC) measurements.
Such calculations, supported by SAXS data, evidenced a multimeric
assembly of up to eight monomers, formed by a linear arrangement of
concatenated dimers. The calculated structure unveiled two solvent-acces-
sible BBXB motifs in each monomer, which become lengthwise-aligned in
the oligomer, thus forming sites on which GAG chains would bind.
Moreover, this arrangement offers an explanation to the promotion of
chemokine oligomerisation exerted by GAG binding, since electrostatic
repulsion of positively-charged residues in the binding site would keep the
oligomer size small unless they are neutralised by negatively-charged
sulphated residues of GAGs. In this context, the characterisation and
binding studies of the GAGs [83] (sulfated glycosaminoglycans) are
206 | Carbohydr. Chem., 2012, 38, 192–214

fundamental from the biological viewpoint because they regulate the
activity of, and bind to, a number of different types of proteins. The analysis
of the heparin pentasaccharide in the free state and in complex with FGF2
receptor indicated the existence of a conformational selection process.
Although there is equilibrium of the 2-O-sulfo-a-L-iduronate ring (IdoA2S)
in the free state between the 1C4 and 2S0 conformers (ratio 60:40), FGF2
receptor selects only the unique twisted-boat 2S0 conformation of this
residue, changing the major conformation of the pentasaccharide in
solution.
Turning back to methodological aspects of STD experiments, the last
years have witnessed a significant expansion of its applicability beyond the
single ligand-receptor system, being used to study protein-ligand interac-
tions in viruses, micelles, or even within whole, living cells. For instance,
STD has been used on novovirus-like particles (VLPs) to study their
interaction with inhibitors targeting the fucose-binding site [84], and on
human embryonic kidney cells to detect the recognition of sweet ligands by
the human sweet receptor [85]. The binding of VLPs of a GII.4 norovirus
strain to host attachment factors has also been investigated by NMR
spectroscopy [86]. STD NMR experiments in conjunction with transfer
NOESY experiments have been employed to elucidate the binding epitopes
and bioactive conformations of synthetic HBGA fragments. The experi-
ments deliver for the first time a systematic overview of the structural
requirements for norovirus attachment to host cell antigens (HBGAs) at
atomic resolution.
Mohan et al. [87] have reported on the exploitation of the 150-cavity in
the active sites of group-1 neuraminidases for the design of new triazole-
containing carbocycles related to oseltamivir. Inhibition studies with virus-
like particles (VLPs) containing the influenza virus neuraminidase-1 (N1)
activity indicate that several candidates are inhibitors, with Ki values in
the 105–108 M range. In contrast, a known candidate that preserves
the free amino group and a new candidate containing a guanidine function
are better inhibitors, with Ki values of 1.5 109 and 4.6 1010 M,
respectively. The most active inhibitor of the N1 enzyme in the triazole
series was selective for the N1 class and showed significantly less inhibition
(Ki=2.6 mM vs 0.07 mM) of the free influenza virus neuraminidase-2 (N2).
STD-NMR spectroscopic studies with this compound and the VLPs show
that the entire molecule forms contacts with residues in the active site.
To overcome the challenge posed by the tendency of many cell lines
to aggregate and to settle, a methodology combining STD with high reso-
lution-magic angle spinning (HR-MAS) has been developed and applied
to detect the interaction of the sugar co-transporter SGTL1 with its
ligands [88].
Among the protein-detected methods, those relying on the receptor epi-
tope mapping through the study of resonance shifting in ‘‘finger-print’’
spectra (such as 1H-15N HSQC and 1H-13C HSQC) are of paramount
importance. 1H, 13C and 15N NMR assignments of several carbohydrate-
binding proteins have been recently reported, including those of galectin-1
[89], galectin-7 [90] and a bacterial D-allose binding protein [91], both in its
free form and in complex with the sugar. It is foreseen that these works
Carbohydr. Chem., 2012, 38, 192–214 | 207

should pave the way for the study of the interactions between these mole-
cular receptors and their carbohydrate ligands in the near future.
4.2 Carbohydrate-carbohydrate and carbohydrate-nucleic acid interactions

NMR experiments using deuterium isotope shifts can be use for analysis of
Lewis X–Lewis X interactions under 100% water conditions, similar to the
biological environment [92]. The authors focused on fast exchange O–H
groups of the Lewis X glycan because they are expected to be directly
involved in stabilizing tertiary structures and interactions with other
molecules by hydrogen bonding and/or coordinating the metallic ions.
Lewis X–Lewis X interactions were analyzed through measurement of
proton-exchange rates under physiological conditions. The technique is able
to quantify the proton exchange of hydroxyl groups on glycans in 100%
water, in the ten to hundred milliseconds range. Also, it was revealed that
Ca2 þ slows down the rates of H/D exchange of several hydroxyl protons of
Lewis X. These groups must be involved in the Lewis X–Lewis X interac-
tion. Not only hydroxyl protons could be observed but also precisely
evaluated their dynamics through measurement of the small difference in
their exchange rates. This simple 13C-NMR strategy is suitable for ana-
lyzing extremely weak interactions including those between carbohydrates,
and permits identifying the hydroxyl protons involved in hydrogen bonding
and metal coordination.
Regarding amino glycosides, the structure of the complex between
apramycin and Cu(II) has been studied by NMR. The paramagnetic
relaxation enhancements provided distance restraints that were employed
for molecular mechanics calculations to deduce the structure of the complex
[93]. On the other hand, the participation of CH/p stacking interactions in
the recognition of amino glycosides by RNA has also been reported to be
crucial. Thus, by using kanamycin derivatives as simple synthetic models,
the role played by CH/p contacts as well as electrostatic forces and solvation
of the charged groups have been examined by NMR experiments, and
molecular dynamics and quantum mechanics calculations. It has been
shown that the recognition phenomenon is extremely dependent on the
protonation state of the interacting amino sugars [94].
The interaction of sugar fragments with DNA has also been studied by
means of NMR. A series of glyco-oligoamides, with different sugars
attached to bi-pyrrol moieties, have been synthesised. The sugar residues
modulated the interaction with DNA fragments and contribute to the
selectivity of the binding. The 3D structure and binding features of the
b-galactose-containing oligoamide with a DNA model sequence was eluci-
dated by standard NMR and modelling methods. The bound galacto-
oligoamide adopts a crescent shaped hairpin conformation, and is located
within the minor groove of the DNA fragment, at its central AATT region
[95]. Also, the sugar moiety of the anthracyclines seems to be critical for
their antitumor activity through its binding to the minor groove of DNA. In
this context, the complexes between abarubicin (Men10755) and daunor-
ubicin (Men10749), two anthracycline disaccharide derivatives, with double
helix oligonucleotides d(CGTACG)2 and d(CGATCG)2 have also been
elucidated by NMR [96].
208 | Carbohydr. Chem., 2012, 38, 192–214

The importance of the interaction of carbohydrates with DNA has also
been studied on carbohydrate-DNA conjugate models, where the oligonu-
cleotide is capped by a mono or a disaccharide attached to the 5 0 end of the
DNA strand. Chemical shift variations at the DNA between the species
conjugated and non conjugated to the sugar, together with NOEs analysis
permitted the authors to guess that the polar sugar moieties may stack onto
DNA duplexes and stabilise sequences with terminal C-G or G-C base pairs,
although with low affinity [97].
4.3 Interactions of carbohydrates with artificial receptors

The development of artificial receptors for carbohydrates has received
increasing attention in recent years, being NMR one of the main tools to
study the binding process. Very different chemical architectures have been
described in order to mimic the molecular recognition process carried out by
the carbohydrate binding proteins in Nature. They share however, common
features: on one hand, the occurrence of groups capable of participating in
hydrogen bonds and, on the other hand, the presence of aromatic moieties
being able to form CH-p stacking interactions.
Therefore, NMR has been employed in order to shed light on the
structural details as well as to assess binding affinities of the recognition
processes. The group of Davis at Bristol has found out that a cage-based
receptor, that had demonstrated to recognise glucose in water, was actually
a better receptor for b-GlcNAc [98]. The NOESY spectrum of the complex
between the synthetic receptor and GlcNAcbOMe allowed identifying an
important number of intermolecular NOEs, providing key information for
describing the interactions responsible for the binding process at the atomic
level. Thus, the recognition process takes place with similar affinity and
selectivity to that occurring in Nature, also employing a similar combina-
tion of CH-p and hydrogen bonding between b-GlcNAc and the receptor.
Indeed, these similarities have led to entitle this type of receptors as ‘syn-
thetic lectins’ [99]. Further studies using different O-glycopeptide fragments
permitted to unravel the different contributions for O-b-GlcNAc-glyco-
peptide recognition by these synthetic receptors [100]. Synthetic modifica-
tions on the receptor with further introduction of substituents at the
biphenyl groups led to a modulation in the affinities and selectivities. In
particular, the attachment of alkoxy groups increased affinity for b-glucosyl
groups [101], while the introduction of electron donor or withdrawing
substituents modified the binding affinities, but to a lesser extent than that
expected, evidencing the delicate interplay of different factors for the
recognition of carbohydrates in water [102].
Other scaffolds for constructing synthetic carbohydrate receptors have
also been investigated. The triethyl benzene structure differently substituted
has been widely explored for sugar recognition in polar organic solvents
[103]. Selective recognition of b-mannosides has been achieved by a hex-
aamino dipyrrolic derivative of the triethyl benzene scaffold containing a
macrocycle in the structure [104, 105]. Moreover, this new receptor showed
chiral discrimination, since the two possible enantiomeric forms of the
receptor displayed different affinities for a- and b-mannosides. The 3D
structure of all of the complexes was elucidated by a combination of NMR
Carbohydr. Chem., 2012, 38, 192–214 | 209

methods with molecular modelling permitting to explain the observed
selectivities [106].
The effect of the substitution of triethyl benzene core-based artificial
receptors on their recognition properties has also been explored. NMR
chemical shift perturbation methods, with either direct or reverse titrations
were employed to monitor the affinity and selectivity towards mono-
saccharide derivatives in organic solvents. For instance, the substitution of
the core scaffold with 8-hydroxyquinoline groups yielded receptors that
showed preferred binding toward b-configurations over the a analogues.
One of these newly synthesised receptors bearing two 8-hydroxyquinoline
fragments was found to effectively bind b-galactosides, while the one
bearing just one 8-hydroxyquinoline moiety showed preference for the b-
glucoside analogue [107]. In contrast, decoration of the same scaffold with
phenantroline groups yielded receptors with binding preference for a-versus
b-anomers [108]. On the other hand, the benzene core scaffold has also been
decorated with isopropyl- and isobutyl-amino functionalities providing
good affinities [109], while the use of larger aromatic surface to host dis-
accharides has also been explored, by employing dimesitylmethane groups
[110].
A new group of artificial receptors has been reported, based on a com-
pletely different scaffold. Porphyrin was chosen for providing the aromatic
surface, to which various chemical functionalities were attached. In this
case, NMR titrations were further supported by other techniques (UV
titrations and CD measurements) to demonstrate that these receptors are
able to recognise pyranose and furanose rings with affinities in the mM
range in methanol/chloroform mixtures, using a combination of CH-p
interactions provided by the porphyrin moiety and of hydrogen bonds
provided by the polar chemical groups attached to it [111].
Acknowledgments
We thank the Ministry of Economy and Competitiveness (MINECO) of
Spain for financial support (Grant CTQ2009-08536). AA, AC, AB, PB,
MCFA, and FM thank MINECO (Juan de la Cierva), MINECO (Ramón y
Cajal), MINECO (FPI), EU (GlycoHit), CSIC (JAE-DOC), and FCT
institutions for funding, respectively.
References
1 X. Hu, W. Zhang, I. Carmichael and A. S. Serianni, J. Am. Chem. Soc., 2010,
132, 4641.
2 X. Hu, I. Carmichael and A. S. Serianni, J. Org. Chem., 2010, 75, 4899.
3 B. Yu, H. van Ingen, S. Vivekanandan, C. Rademacher, S. E. Norris and D. I.
Freedberg, Journal of Magnetic Resonance, 2012, 215, 10.
4 M. Hricovini, J. Phys. Chem. B, 2011, 115, 1503.
5 E. Säwén, B. Stevensson, J. Östervall, A. Maliniak and G. Widmalm, J. Phys.
Chem. B, 2011, 115, 7109.
6 J. Xia, C. J. Margulis and D. A. Case, J. Am. Chem. Soc., 2011, 133, 15252.
7 M. Erdèlyi, E. d’Auvergne, A. Navarro-Vázquez, A. Leonov and C.
Griesinger, Chem. Eur. J., 2011, 17, 9668.
210 | Carbohydr. Chem., 2012, 38, 192–214

8 a) A. Mallagaray, G. Domı́nguez, J. Jiménez-Barbero and J. Pérez-Castells,
Chem Commun (Camb), 2011, 47, 7179. b) S. Yamamoto, T. Yamaguchi, M.
Erdélyi, C. Griesinger and K. Kato, Chemistry, 2011, 17, 9280.
9 A. Gallinger, T. Biet, L. Pellerin and T. Peters, Angew. Chem. Int. Ed., 2011,
50, 11672.
10 A. Imberty and S. Perez, Chem. Rev., 2000, 100, 4567.
11 B. Mulloy, M. J. Forster, C. Jones and D. B. Davies, Biochem. J., 1993, 293, 849.
12 M. Sattelle, S. U. Hansen, J. Gardiner and A. Almond, J. Am. Chem. Soc.,
2010, 132, 13132.
13 M. L. DeMarco and R. J. Woods, Glycobiology, 2008, 18, 426.
14 H. Kogelberg, D. Solı́s and J. Jiménez-Barbero, Curr. Opin. Struct. Biol., 2003,
13, 646.
15 K. N. Kirschner and R. J. Woods, Proc. Natl. Acad. Sci. USA, 2001, 98,
10541.
16 V. Roldós, F. J. Cañada and J. Jiménez-Barbero, ChemBioChem, 2011, 12, 990.
17 F.-X. Theillet, C. Simenel, C. Guerreiro, A. Phalipon, L. A. Mulard and M.
Delepierre, Glycobiology, 2011, 21, 109.
18 U. Olsson, E. SäwØn, R. Stenutz and G. Widmalm, Chem. Eur. J., 2009, 15,
8886.
19 E. Hatcher, E. Sawen, G. Widmalm and A. D. MacKerell, Jr., J. Phys. Chem.
B, 2011, 115, 597.
20 E. Säwén, T. Massad, C. Landersjö, P. Damberg and G. Widmalm, Org.
Biomol. Chem., 2010, 8, 3684.
21 Y. Kamiya, S. Yamamoto, Y. Chiba, Y. Jigami and K. Kato, J Biomol NMR,
2011, 50, 397.
22 E. Säwén, M. U. Roslund, I. Cumpstey and G. Widmalm, Carbohydrate
Research, 2010, 345, 984.
23 B. M. Sattelle, J. Shakeri, I. S. Roberts and A. Almond, Carbohydrate
Research, 2010, 345, 291.
24 K. Hanna, M. Jonsson, R. Pendrill and G. Widmalm, Magn. Reson. Chem.,
2011, 49, 117.
25 P. J. Meloncelli, L. J. West and T. L. Lowary, Carbohydrate Research, 346,
1406.
26 T. Fujimoto, T. Kato, Y. Usui, O. Kamo, K. Furihata, K. Tsubono, T. Kato,
T. Machinami and M. Tashiro, Carbohydrate Research, 2011, 346, 1991.
27 M. Zerbetto, A. Polimeno, D. Kotsyubynskyy, L. Ghalebani, J. Kowalewski,
E. Meirovitch, U. Olsson and G. Widmalm, J Chem Phys., 2009, 131, 234501.
28 C. D. Blundell, I. S. Roberts, J. K. Sheehan and A. Almond, J Mol Microbiol
Biotechnol, 2009, 17, 71.
29 V. Gargiulo, M. A. Morando, A. Silipo, A. Nurisso, S. Pérez, A. Imberty, F. J.
Cañada, M. Parrilli, J. Jiménez-Barbero and C.D. Castro, Glycobiology, 2010,
20, 1208.
30 E. Säwén, E. Huttunen, X. Zhang, Z. Yang and G. Widmalm, J Biomol NMR,
2010, 47, 125.
31 S. E. Norris, J. Landstrom, A. Weintraub, T. E. Bull, G. Widmalm and D. I.
Freedberg, Biopolymers, 97, 145.
32 a) M. Lundborg and G. Widmalm, Anal. Chem., 2011, 83, 1514. b) M.
Lundborg, C. Fontana and G. Widmalm, Biomacromolecules, 2011, 12, 3851.
c) M. U. Roslund, E. Säwén, J. Landström, J. Rönnols, K. Hanna, M.
Jonsson, M. Lundborg, M. V. Svensson and G. Widmalm, Carbohydrate
Research, 346, 1311.
33 A. W. Barb, A. J. Borgert, M. Liu, G. Berany and D. Live, Methods Enzymol.,
2010, 478, 365.
Carbohydr. Chem., 2012, 38, 192–214 | 211

34 A. Varki, Essentials of Glycobiology, Cold Spring Harbor Laboratory Press:
Cold Spring Harbor, NY, USA, 2009.
35 K. G. Ten Hagen, T. A. Fritz and L. A. Tabak, Glycobiology, 2003, 13, 1.
36 M. A. Tarp and H. Clausen, Biochim. Biophys. Acta, 2008, 1780, 546.
37 B. Meyer and H. Möller, Top. Curr. Chem., 2007, 267, 187.
38 D. M. Coltart, A. K. Royyuru, L. J. Williams, P. W. Glunz, D. Sames, S. D.
Kuduk, J. B. Schwarz, X.-T. Chen, S. J. Danishefsky and D. H. Live, J. Am.
Chem. Soc., 2002, 124, 9833.
39 F. Corzana, J. H. Busto, M. Garcia de Luis, J. Jiménez-Barbero, A. Avenoza
and J. M. Peregrina, Chem. Eur. J., 2009, 15, 3863.
40 F. Corzana, J. H. Busto, F. Marcelo, J. L. Asensio, S. Martı́n-Santamarı́a, M.
Garcia de Luis, J. Jiménez-Barbero, A. Avenoza and J. M. Peregrina, Chem.
Eur. J, 2011, 17, 3105.
41 R. Hashimoto, N. Fujitani, Y. Takegawa, M. Kurogochi, T. Matsushita, K.
Naruchi, N. Ohyabu, H. Hinou, X. D. Gao, N. Manri, H. Satake, A. Kaneko,
T. Sakamoto and S-I. Nishimura, Chem. Eur. J., 2011, 17, 2393.
42 G. Arsequell, M. Rosa, C. Mayato, R. L. Dorta, V. Gonzalez-Nunez,
K. Barreto-Valer, F. Marcelo, L. P. Calle, J. T. Vázquez, R. E. Rodrı́guez,
J. Jiménez-Barbero and G. Valencia, Org. Biomol. Chem., 2011, 9, 6133.
43 Y. Yamaguchi and K. Kato, Methods Enzymol., 2010, 478, 305.
44 K. Kato, Y. Yamaguchi and Y. Arata, Prog. Nucl. Magn. Reson. Spectrosc.,
2010, 56, 346.
45 A. W. Barb and J. H. Prestegard, Nat. Chem. Biol., 2011, 7, 147.
46 K.-F. Mo, T. Fang, S. H. Stalnaker, P. S. Kirby, M. Liu, L. Wells, M. Pierce,
D. H. Live and G.-J. Boons, J. Am. Chem. Soc., 2011, 133, 14418.
47 F. Squeglia, R. Marchetti, A. Ruggiero, R. Lanzetta, D. Marasco, J. Dworkin,
M. Petoukhov, A. Molinaro, R. Berisio and A. Silipo, J. Am. Chem. Soc.,
2011, 133, 20676.
48 M. M. Mackeen, A. Almond, M. Deschamps, I. Cumpstey, A. J. Fairbanks,
C. Tsang, P. M. Rudd, T. D. Butters, R. A. Dwek and M. R. Wormald,
J. Mol. Biol., 2009, 387, 335.
49 a) Y. Yamaguchi and K. Kato, Methods in Enzymology, 478. b) K. Kato,
Y. Yamaguchi and Y. Arata, Progress in Nuclear Magnetic Resonance Spec-
troscopy, 2010, 56, 346.
50 J. L. Magnani and B. Ernst, Discover. Med., 2009, 43, 247.
51 V. Roldos, F. J. Canada and J. Jimenez-Barbero, ChemBioChem, 2011, 12,
990.
52 M. von Itzstein, Nat. Rev. Drug Discovery, 2007, 6, 967.
53 B. J. Gross, J. G. Swoboda and S. Walker, J. Am. Chem. Soc., 2008, 130, 440.
54 F. Marcelo, Y. He, S. A. Yuzwa, L. Nieto, J. Jiménez-Barbero, M. Sollogoub,
D. J. Vocadlo, G. D. Davies and Y. Bleériot, J. Am. Chem. Soc., 2009, 131,
5390.
55 T. Kuntothom, M. Raab, I. Tvaroska, S. Fort, S. Pengthaisong, J. Cañada,
L. Calle, J. Jiménez-Barbero, J. R. K. Cairns and M. Hrmova, Biochemistry,
2010, 49, 8779.
56 P. J. Hajduk and J. Greer, Nat. Rev. Drug Discovery, 2007, 6, 211.
57 R. Caraballo, H. Dong, J. P. Ribeiro, J. Jiménez-Barbero and O. Ramström,
Angew. Chem. Int. Ed., 2010, 49, 589.
58 R. Caraballo, M. Sakulsombat and O. Ramström, ChemBioChem, 2010, 11,
1600.
59 J. J. Reina, S. Sattin, D. Invernizzi, S. Mari, L. Martinez-Prats, G. Tabarani,
F. Fieschi, R. Delgado, P. M. Nieto, J. Rojo and A. Bernardi, ChemMedChem,
2007, 2, 1030.
212 | Carbohydr. Chem., 2012, 38, 192–214

60 J. J. Reina, I. Diaz, P. M. Nieto, N. E. Campillo, J. A. Paez, G. Tabarani,
F. Fieschi and J. Rojo, Org. Biomol. Chem., 2008, 6, 2743.
61 G. Timpano, G. Tabarani, M. Anderluh, D. Invernizzi, F. Vasile, D. Potenza,
P. M. Nieto, J. Rojo, F. Fieschi and A. Bernardi, ChemBioChem, 2008, 9, 1921.
62 C. Guzzi, J. Angulo, F. Doro, J. J. Reina, M. Thépaut, F. Fieschi, A. Bernardi,
J. Rojo and P. M. Nieto, Org. Biomol. Chem., 2011, 9, 7705.
63 M. Morando, Y. Yao, S. Martı́n-Santamarı́a, Z. Zhu, T. Xu, F. Javier
Cañada, Y. Zhang and J. Jiménez-Barbero, Chem. Eur. J., 2010, 16, 4239.
64 Y. M. Chabre and R. Roy, in The Sugar Code: Fundamentals of Glyco-
sciences (Ed.: H.-J. Gabius), Wiley-VCH, Weinheim, 2009, pp. 53 –70.
65 Y. M. Chabre, D. Giguère, B. Blanchard, J. Rodrigue, S. Rocheleau, M.
Neault, S. Rauthu, A. Papadopoulos, A. A. Arnold, A Imberty and R. Roy,
Chem. Eur. J., 2011, 17, 6545.
66 D. Schwizer, J. T. Patton, B. Cutting, M. Smiesko, B. Wagner, A. Kato, C.
Weckerle, F. P. C. Binder, S. Rabbani, O. Schwardt, J. L. Magnani and Beat
Ernst, Chem. Eur. J., 2012, 18, 1342.
67 a) M. G. Szczepina, D. W. Bleile, J. Müllegger, A. R. Lewis and B. M. Pinto,
Chem. Eur. J., 2011, 17, 11438. b) M. G. Szczepina, D. W. Bleile and B. M.
Pinto, Chem. Eur. J., 2011, 17, 11446.
68 H.-C. Siebert, J. Jiménez-Barbero, S. André, H. Kaltner and H.-J. Gabius,
Methods Enzymol, 2003, 362, 417.
69 H. C. Siebert, S. André, Y. Lu, M. Frank, H. Kaltner, J. A. van Kuik, E. Y.
Korchagina, N. Bovin, E. Tajkhorshid, R. Kaptein, J. F. Vliegenthart, C. W.
von der Lieth, J. Jiménez-Barbero, J. Kopitz and H. J. Gabius, Biochemistry,
2003, 42, 14762.
70 V. Roldos, F. J. Canada and J. Jimenez-Barbero, Chembiochem, 2011, 12, 990.
71 J. Angulo, P. M. Enriquez-Navas and P. M. Nieto, Chemistry, 2010, 16, 7803.
72 M. Marradi, P. Di Gianvincenzo, P. M. Enriquez-Navas, O. M. Martinez-
Avila, F. Chiodo, E. Yuste, J. Angulo and S. Penades, J. Mol. Biol., 2011, 410,
798.
73 M. Plum, Y. Michel, K. Wallach, T. Raiber, S. Blank, F. I. Bantleon, A.
Diethers, K. Greunke, I. Braren, T. Hackl, B. Meyer and E. Spillner, J. Biol.
Chem., 2011, 286, 43103.
74 M. C. Miller, J. P. Ribeiro, V. Roldos, S. Martin-Santamaria, F. J. Canada,
I. A. Nesmelova, S. Andre, M. Pang, A. A. Klyosov, L. G. Baum, J. Jimenez-
Barbero, H. J. Gabius and K. H. Mayo, Glycobiology, 2011, 21, 1627.
75 L. M. Koharudin and A. M. Gronenborn, Structure, 2011, 19, 1170.
76 S. Liu, L. Meng, K. W. Moremen and J. H. Prestegard, Biochemistry, 2009, 48,
11211.
77 S. Liu, A. Venot, L. Meng, F. Tian, K. W. Moremen, G. J. Boons and J. H.
Prestegard, Chem. Biol., 2007, 14, 409.
78 K. Schaefer, J. Albers, N. Sindhuwinata, T. Peters and B. Meyer, ChemBio-
Chem, 2012, 13, 443.
79 C. Rademacher, J. Landström, N. Sindhuwinata, M. M. Palcic, G. Widmalm
and T. Peters, Glycoconj. J., 2010, 27, 349.
80 a) M. G. Szczepina, R. B. Zheng, G. C. Completo, T. L. Lowary and B. M.
Pinto, ChemBioChem, 2009, 10, 2052. b) M. G. Szczepina, R. B. Zheng, G. C.
Completo, T. L. Lowary and B. M. Pinto, Bioorganic & Medicinal Chemistry,
2010, 18, 5123.
81 A. L. Jansma, J. P. Kirkpatrick, A. R. Hsu, T. M. Handel and D. Nietlispach,
J. Biol. Chem., 2010, 285, 14424.
82 X. Wang, C. Watson, J. S. Sharp, T. M. Handel and J. H. Prestegard,
Structure, 2011, 19, 1138.
Carbohydr. Chem., 2012, 38, 192–214 | 213

83 L. Nieto, A. Canales, G. Giménez-Gallego, P.M. Nieto and J. Jiménez-
Barbero, Chem. Eur. J., 2011, 17, 11204.
84 C. Rademacher, J. Guiard, P. I. Kitov, B. Fiege, K. P. Dalton, F. Parra, D. R.
Bundle and T. Peters, Chemistry, 2011, 17, 7442.
85 F. M. Assadi-Porter, M. Tonelli, E. Maillet, K. Hallenga, O. Benard, M. Max
and J. L. Markley, J. Am. Chem. Soc., 2008, 130, 7212.
86 B. Fiege, C. Rademacher, J. Cartmell, P. I. Kitov, F. Parra and T. Peters,
Angew. Chem. Int. Ed., 2012, 51, 928.
87 S. Mohan, S. McAtamney, T. Haselhorst, M. von Itzstein and B. M. Pinto,
J. Med. Chem., 2010, 53, 7377.
88 C. Airoldi, S. Giovannardi, B. La Ferla, J. Jimenez-Barbero and F. Nicotra,
Chemistry, 2011, 17, 13395.
89 I. V. Nesmelova, M. Pang, L. G. Baum and K. H. Mayo, Biomol. NMR
Assign., 2008, 2, 203.
90 I. V. Nesmelova, M. A. Berbis, M. C. Miller, F. J. Canada, S. Andre, J.
Jimenez-Barbero, H. J. Gabius and K. H. Mayo, Biomol. NMR Assign., 2011.
91 D. Castano and O. Millet, Biomol. NMR Assign., 2011, 5, 31–34.
92 S. Hanashima, K. Kato and Y. Yamaguchi, Chem. Commun., 2011, 47, 10800.
93 D. Balenci, G. Bonechi, N. D’Amelio, E. Gaggelli, N. Gaggelli, E. Molteni, G.
Valensin, W. Szczepanik, M. Dziuba, G. Swiecicki and M. Jezowska-Bojczuk,
Dalton Trans., 2009, 7, 1123.
94 T. Vacas, F. Corzana, G. Jimeénez-Osés, C. González, A. M. Gómez, A.
Bastida, J. Revuelta and J. L. Asensio, J. Am. Chem. Soc., 2010, 132, 12074.
95 P. Peñalver, F. Marcelo, J. Jiménez-Barbero and C. Vicent, Chem. Eur. J.,
2011, 17, 4561.
96 S. Mazzini, L. Scaglioni, F. Animati and R. Mondelli, Bioorg. Med. Chem.,
2010, 18, 1497.
97 R. Lucas, I. Góomez-Pinto, A. Aviñó, J. J. Reina, R. Eritja, C. Gonzáalez and
J. C. Morales, J. Am. Chem. Soc., 2011, 133, 1909.
98 Y. Ferrand, E. Klein, N. P. Barrwell, M. C. Crumps, J. Jiménez-Barbero,
C. Vicent, G.-J. Boons, S. Ingale and A. P. Davis, Angew. Chem. Int. ed., 2009,
48, 1775.
99 S. Kubik, Angew. Chem. Int. Ed., 2009, 48, 1722.
100 F. Corzana, A. Fernández-Tejada, J. H. Busto, G. Joshi, A. P. Davis,
J. Jiménez-Barbero, A. Avenoza and J. M. Perengrina, ChemBioChem, 2011,
12, 110.
101 N. P. Barrwell, M. C. Crumps and A. P. Davis, Angew. Chem, Int. ed., 2009,
48, 7673.
102 N. P. Barrwell, J. Org. Chem., 2011, 76, 6548.
103 M. Cacciarini, C. Nativi, M. Norcini, S. Staderini, O. Francesconi and
S. Roelens, Org. Biomol. Chem., 2011, 9, 1085.
104 A. Ardá, C. Venturi, C. Nativi, O. Francesconi, G. Gabrielli, F. J. Cañada,
J. Jiménez-Barbero and S. Roelens, Chemistry, 2010, 16, 414.
105 C. Nativi, O. Francesconi, G. Gabrielli, A. Vacca and S. Roelens, Chemistry,
2011, 17, 4814.
106 A. Ardá, F. J. Cañada, C. Nativi, O. Francesconi, G. Gabrielli, A. Ienco,
J. Jiménez-Barbero and S. Roelens, Chemistry, 2011, 17, 4821.
107 M. Mazik and C. Geffert, Org. Biomol. Chem., 2011, 9, 2319.
108 M. Mazik, A. Hatmann and P. G. Jones, Chemistry Eur. J., 2009, 15, 9147.
109 M. Mazik and C. Sonnenberg, J. Org. Chem., 2010, 75, 6416.
110 M. Mazik and A. C. Buthe, Org. Biomol. Chem., 2009, 7, 2063.
111 J.-D. Lee, Y.-H. Kim and J.-I. Hong, J. Org. Chem., 2010, 75, 7588.
214 | Carbohydr. Chem., 2012, 38, 192–214

Glycosidase inhibitors: versatile tools
in glycobiology
Óscar López,* Penélope Merino-Montiel, Sergio Martos and
Alejandro González-Benjumea
DOI: 10.1039/9781849734769-00215
Imino sugars, despite being firstly studied as inhibitors more than forty
years ago, are still a relevant target in Medicinal and Bioorganic Chemistry,
because of their numerous and substantial biological activities. This review
covers the most significant reports concerning the synthesis and evaluation
of imino sugars as glycosidase inhibitors in the period 2007–2012. The aim is
to present the synthesis of new different families and their biological assays,
ranging from in vitro tests to animal models.
1 Introduction
Ubiquitous carbohydrate-processing enzymes like glycosidases, glycosyl
transferases, and glycogen phosphorylase participate in pivotal biological
processes in living organisms such as carbohydrate catabolism, cell wall
formation, or molecular recognition, among others.1
Such enzymes constitute a therapeutic target of great interest in Medicinal
Chemistry, as their anomalous functioning is correlated to the development
of numerous diseases, like lysosomal storage disorders, diabetes, or cancer;
furthermore they also participate in carbohydrate-carbohydrate recognition
processes associated to some microbial infections.1
Classical description of glycosidases divides these enzymes into two main
categories: retaining or inverting glycosidases, depending on the catalytic
activity exerted by the carboxyl/carboxylate moieties at the active site.
More recently some other mechanisms have been proposed2 for small
subsets of hydrolases, like a substrate-assisted catalytic mechanism for some
hexosaminidases, or a tyrosine residue acting as the catalytic nucleophile for
some sialidases and neuraminidases.
In this context, potent and selective enzymatic inhibitors, either synthetic
or naturally-occurring, have been postulated as potential pharmacological
agents. Among the different families of glycosidase inhibitors, those with a
nitrogen atom replacing either the endocyclic oxygen (imino sugars) or a
carbon atom (aza sugars) have received great attention. Many of the classical
glycosidase inhibitors fall into the following categories: polyhydroxylated
pyrrolidines, piperidines or bicyclic templates such as indolizidines, pyrro-
lizidines or nortropanes,1 but the number of families has grown up intensively
in the last decade. These compounds might act as transition state analogues
in terms of geometry and charge; nevertheless, this assumption is highly
discussed. Withers claimed3 that the sp2-hybridized imino sugars can be
good transition state mimics, whereas sp3-derivatives, despite exhibiting in
many cases high affinity to the enzyme, do not fulfill this definition.
Departamento de Quı´mica Orgánica, Facultad de Quı´mica, Universidad de Sevilla, c/Profesor

Garcıá González 1, E-41012, Seville, Spain. E-mail: osc-lopez@us.es
Carbohydr. Chem., 2012, 38, 215–262 | 215

Glycosidase inhibitors are also suitable as chemical probes for mechan-
istic or structural studies concerning glycosidases. For instance, Pieter’s
group reported4 the synthesis of an amide derived from deoxynojirimycin
for specific labeling of lysosomal glucocerebrosidase; this imino sugar
exhibited high affinity in the presence of cell lysate. This provides a method
for studying structural requirements in the enzyme active site for potent
enzyme-inhibitor interactions.
Davies and co-workers made an exceptional contribution5 to mechanistic
studies of glycosidase inhibition by providing a panel of experimental data on
the pH dependence of 18 glycosidase inhibitors binding to TMGHa b-gluco-
sidase. Furthermore, a collection of atomic resolution X-ray crystallography
of enzyme-inhibitors complexes6 and titration calorimetry data were also
included. Bols and co-workers studied7 the inhibition profile of 13C or 15N-
labeled azafagomine, a cyclic polyhydroxylated hydrazine. For this purpose,
magic angle spinning (MAS) solid NMR of inhibitor complexes was used,
deducing that title compound bound to b-glucosidase in its protonated state,
whereas non-protonated when interacting with a-glucosidase or glucoamylase.
The aim of this review is to carry out a survey through the most striking
aspects concerning glycosidase inhibitors in the period 2007-2012, including
a coverage of both, synthetic and biological aspects.
2 Polyhydroxylated pyrrolidines
There is a huge number of imino sugar derivatives bearing a pyrrolidine
template, many of them exhibiting inhibition constants in the sub-
micromolar range, probably by mimicking the exo-protonated glycoside in
the course of the hydrolysis. Usually, pyrrolidine-based inhibitors show
reduced selectivity when compared with six-membered ring imino sugars,
due to a major conformational flexibility. Like piperidine-based imino
sugars, despite the plethora of different structures reported so far, intense
research is still kept in the search for better prototypes and mechanistic
studies. In this context, the first X-ray crystal structure of a pyrrolidine
imino sugar complexed with a b-glycosidase was reported in 2007.8
In a recent work,9 L-DMDP 1 and L-homoDMDP 2 (Fig. 1), the
L-enantiomers of DMDP and homoDMDP, have been synthesized via
sugar-derived cyclic nitrones as well as 3-deoxy-3-fluoro derivatives.
Inhibition assays confirmed that DMDP and homoDMDP are potent
a-glycosidase inhibitors, whereas C3-fluorinated analogues showed no
inhibition against any of the glycosidases tested; these results proved that
C3 hydroxyl group is crucial in the interaction with these enzymes. Intro-
ducing a branched carbon chain to the pyrrolidine core is a novel approach
to obtain glycosidase inhibitors; in this context, isoDAB 17 and its enan-
tiomer isoLAB are positional isomers of DAB and LAB and bear a bran-
ched hydroxymethyl group at C2. Their syntheses from D-ribose (Scheme 1)
and D-tagatose, together with their biological activities have been studied by
Best and co-workers.10 IsoDAB was found to be the first example of a
carbon-branched pyrrolidine acting as a potent and specific inhibitor of
a-glucosidases; on the contrary, isoLAB did not inhibit significantly any of
the glycosidases tested.
216 | Carbohydr. Chem., 2012, 38, 215–262

HO OH HO OH HO OH
HO
N N N
H H H
HO OH OH OH
1 HO 2 3
L-DMDP L-homoDMDP
Ki 4 μM
IC50 0.83 μM (α-glucosidase) IC50 5.5 μM (α-glucosidase) (α-L-fucosidase)
HO OH
HO OH HO OH
Me
N
N N H
H H NH
HO HO Ar
4 5 6
Ki 8 μM IC50 9 μM Ar = diphenyl, Ki 0.040 μM
(α-L-fucosidase) (α-L-fucosidase) (α-L-fucosidase)
HO OH
H HO
N
H HOH2C HCl
N N
Me
N
H
N
HO NHCOCH3
HO OH
8 LABNAc
7 9
Ki 0,095 μM
Ki 0.080 μM (β-N-acetyl-D-hexosaminidase Ki 9.8 μM
(α-L-fucosidase)
from bovine kidney) (α-L-fucosidase)
OH
H CH2OH
N (CH2)5
HOH2C H·CF3CO2H
N
OH
HO
HO O
O 10 O OH
IC50 8 μM OH
HO (α-galactosidase) 11
O HO Ki 3.0 μM (Arb93A)
OH
HO CH2OH
OH
Fig. 1
Izquierdo and co-workers reported11 a novel synthesis of known trihy-

droxy methylpyrrolidines 3–5 from orthogonally protected pyrrolidines;
title compounds incorporated a methyl moiety that might resemble L-fuco
configuration, and in fact are potent a-L-fucosidase inhibitors with Ki values
in the micromolar range (4–9 mM) (Fig. 1).
An approach for improving the activity against a-L-fucosidase is the
incorporation of aryl or aminoaryl moieties to the known inhibitor meso-
pyrrolidine-3,4-diol. Among these compounds, remarkable examples are 6
and 7, with potency in the submicromolar range.12
Carbohydr. Chem., 2012, 38, 215–262 | 217

OH
HOH2C O HOH2C CH2OH O

OH NaIO4 HO
1) CH2O, K2CO3 CH2OH
2) NaBH4 H2O CH2OH

O O
O O O O
12 13 14
H
N O O
HO HO
H2, Pd/C Dowex (H+)
HOH2C H2O/AcOH CH2N3 H2O/dioxane CH2N3
HO OH HO OH O O
17
isoDAB 16 15
Ki 4 μM (α-glucosidase)
Scheme 1
OMs
BnO OBn
BnO OBn BnO

CH3MgBr
BnO
OMs toluene N
N
BnO
BnO
BnO CN
CH3 CH3
18 19 20
Scheme 2
Compounds incorporating an acetamido group on pyrrolidines were also

designed as potential hexosaminidase inhibitors; Rountree et al. accom-
plished13 the synthesis of 2-acetamido-1,4-imino-1,2,4-trideoxy-L-arabinitol
(LABNAc, 8) in 11 steps starting from D-lyxonolactone. This compound
was found to be a potent non-competitive inhibitor of b-N-acetylhex-
osaminidases, especially from bovine kidney, with Ki of 95 nM. Its N-butyl
derivative was also prepared and tested, showing a similar potency.
Behr et al. reported14 the synthesis of ketimine-type imino sugars
analogues of nectrisine, an unsaturated 5-membered cyclic imino sugar with
an a-glucosidase inhibition even higher than nojirimycin. The key step is
a tandem addition-cyclization reaction of a Grignard reagent to o-metha-
nesulfonylglyconitrile 18 (Scheme 2). For instance, imino sugar 9 (Fig. 1),
obtained using this methodology, showed a good inhibition (Ki=9.8 mM)
against a-L-fucosidase via a mixed mechanism.
Some imino sugars incorporated into oligosaccharide architectures
have been found in nature; for instance, a-glucoside 10 (Fig. 1) bearing
an unsaturated side chain was isolated from the roots of Adenophora
triphylla.15 Biological activity evaluated against several common
218 | Carbohydr. Chem., 2012, 38, 215–262

glycosidases showed potent and highly specific a-galactosidase inhibition
(IC50=8 mM). Considering such results, a smart strategy for obtaining
potent glycosidase inhibitors could be the design of oligosaccharides con-
taining imino sugar units, like a-L-arabinofuranosylated pyrrolidines.16
Using a selective glycosilation methodology,17 arabinan-like oligosacchar-
ides were prepared, like pseudo-disaccharide 11 (Fig. 1), which was found to
be a good inhibitor of the GH93 exo-1,5-arabinanase Arb93A from
Fusarium graminearum (Ki=3.0 mM).
3 Six-membered ring-based imino sugars

3.1 Piperidines
Since the first isolation of polyhydroxy piperidine derivatives from nature,
such as nojirimycin and 1-deoxynojirimycin, many other synthetic analo-
gues have been developed using various approaches (N-alkylation, sub-
stitution of hydroxyl groups, addition of halogens, etc.) to improve the
spectrum of activity.1,18
One approach for the synthesis of more potent piperidine-based inhibi-
tors is the increase of the lipophilicity by using fatty chains and aromatic
fragments19,20 as appendages; it is postulated that favorable intermolecular
hydrophobic interactions involving lipophilic pockets of the enzyme active
site might lead to more potent and selective inhibitors. For instance, Schitter
et al.21 have modified 1-deoxygalactonojirimycin by N-alkylation with lin-
kers of different lengths bearing terminal aromatic residues. Compounds 21
and 22 (Fig. 2) showed an activity in many cases considerably higher than
parent 1-deoxygalactonojirimycin; these results are especially remarkable
for b-galactosidase from E. coli.
Steiner et al.22 have developed a series of hybrid iminoalditols with
lysine, like 24 (Scheme 3), via a double reductive amination of D-xylo-hexos-
5-ulose (23) with the terminal amino group of a-methyl ester N-Boc-lysine,
followed by dansylation of the secondary amine; the corresponding
N-alkylated derivatives were found to be good inhibitors of b-glucosidase
HO H
N N
N
HO O
HO OH
21 O
HO
N 6 N
Ki (μM) β-Gal (E. Coli) β-Gal (Green H
coffee beans) HO
DNJ 13 0.013
HO OH
21 0.83 2.2 22
22 0.25 7.0
Fig. 2
Carbohydr. Chem., 2012, 38, 215–262 | 219

Dansyl NH OMe
OH HO
HO O N
HO OH HO
HO OH HO OH
23 24
Ki 2.6 μM (β−Glucosidase)
HO
HO
NO2
HO H
HO N
NH
N
5
HO
HO N3
OH
O
25 26
IC50 0.017 and 0.30 μM (α−Glucosidase I and II)
Scheme 3
(Agrobacterium sp.), with a potency higher than parent deoxynojirimycin.

The same authors reported a similar synthesis starting from L-arabino-
hexos-5-ulose.23
Rawlings et al. have also achieved24 the synthesis of potent a-glucosidase
inhibitors based on N-alkylated deoxynojirimycin (Scheme 3); in this con-
text, compound 26 was originally designed as a potential photoaffinity
probe for endoplasmic reticulum a-glucosidases. It was prepared by
reductive amination involving a functionalized aldehyde and the free
nitrogen of deoxynojirimycin. This compound proved to be the hitherto
best inhibitor of a-glucosidase I, with increased potency compared to
N-butyl-deoxynojirimycin, one of the most potent inhibitors reported so far
for this enzyme.
Ardes-Guisot and co-workers25 have developed a series of DNJ-
derivatives by click connection (copper-catalyzed alkyne-azide cycloaddition)
to afford triazole derivatives with adamantane appendages, with different
linker sizes, like 28–30 (Scheme 4). Such compounds behaved as remarkable
inhibitors of different a-glucosidases (rice, maltase and sucrase rat intestinal)
with IC50 values in the low micromolar range (0.030–1.8 mM), all similar or
even better than therapeutical inhibitors like miglitol. Furthermore, triazole
derivatives also showed a high potency against GCase.26
Analogously, Diot et al. described27 the synthesis of some derivatives of
deoxynojirimycin bearing triazolyl moieties with aromatic substituents;
inhibition of GCase exerted by these compounds proved to be strong (IC50
o 5 mM), roughly the similar potency found for N-nonyldeoxynojirimycin.
220 | Carbohydr. Chem., 2012, 38, 215–262

OH
N N N
HO N
OH HO n-1
H OH
N
BnO O
BnO GCase IC50 [µM]
OBn
27 28 n = 4 0.30
29 n = 6 0.46
30 n = 8 0.47
Scheme 4
F
OH OH
F F
HO
NH NH
OH
31 32
Ki 1.2 μM Ki 11.9 μM
(β−Glucosidase) (β−Glucosidase)
Fig. 3
In recent years, another strategy developed for modifying the structure of

imino sugars is the insertion of electron-withdrawing substituents.28 This
idea is based on the fact that strongly electron-withdrawing groups can
greatly affect the pKa of imino sugars, which might have interesting con-
sequences in their mechanism of action against certain glycosidases.28 Li and
co-workers designed29 the preparation of 4-deoxy-4,4-difluoroisofagomine
and analogues from difluoromethylated synthons developed in their research
group. The resulting compounds were not very promising, except difluori-
nated derivative 31 (Fig. 3), which showed a potent and selective b-
glucosidase activity (Ki=1.2 mM).29 Stimulated by these results, some
monofluorinated derivatives based on the isofagomine template have been
synthesized stereoselectively. Imino sugar 3230 (Fig. 3) was obtained, and
found to be a strong inhibitor of b-glucosidase (almond) with a Ki of 11.9 mM.
Another approach that should be mentioned concerning inhibition of
glycosidases is multivalency; this feature has mainly been studied for car-
bohydrate-lectin interactions,31 and currently the number of multivalent
imino sugars as glycosidase inhibitors is scarce (e.g. Figs. 4 and 5). In
principle, considering that glycosidases are monovalent biomolecules, it was
not likely that multivalency could be an important fact concerning such
enzymes. Overkleeft and co-workers accomplished32 the preparation of
dimeric lipophilic deoxynojirimycins of D-gluco and L-ido configurations 37
and 38, respectively, from bis-functionalized adamantanes (Fig. 4). Dimeric
derivatives were in vivo assayed against glycosylceramide synthase (GCS), a
membrane-bound enzyme responsible for the glycosylceramide metabolism,
and the results compared with monomeric deoxynojirimycins 33–36. As can
be seen, a decreased activity was observed for dimeric architectures.
Nevertheless, divalent imino sugars generally kept a comparable potency
against glycosidases (e.g. membrane-bound b-glucosidase 2, a-glucosidase
or sucrose).
Carbohydr. Chem., 2012, 38, 215–262 | 221

OH HO
OH
HO OH
HO N 5O O 5N
N O n
HO OH
HO
OH OH
OH 33 n=1 D-Gluco 37 D-Gluco
34 n=1 L-Ido 38 L-Ido
35 n=2 D-Gluco
36 n=2 L-Ido
GCS in vivo GBA2 in vitro Lysosomal α-glucosidase in vitro Sucrase in vitro
33 0.2 0.001 0.4 0.5
34 0.1 0.03 >100 >100
35 1 0.005 0.8 0.35
36 1 0.006 >100 200
37 20 0.008 0.35 0.4
38 20 0.150 700 200
IC50 in μM
Fig. 4
R
N
N R
N N
R N N
N N R
N
N
O
O N
O O N OH
N N O O O O N
R = HO
N HO
R O O R OH
N
O N N
O
O
N N
O
R N OH N
O
N N
O HO N
O O HO
O O OH
N O 39 Pr
N O N
N R
N O N
O
R
N Enzyme 39 40
N N
N N N N α-Galactosidase (Green coffee beans) N.I. (at 2 mM) 84
R N R
N Isomaltase (baker's yeast) 943 10.5
R α- Mannosidase (Jack bean) 322 0.15
40 Ki in μM
Fig. 5
Recently, copper(I)-catalyzed alkyne-azide cycloaddition (click chemistry

approach) has been used to connect azido-containing N-substituted
deoxynojirimycin to multivalent cores bearing alkynyl residues, like oli-
go(ethyleneglycol) scaffolds,33 fullerenes34 and cyclodextrins;35 some of
these imino sugar derivatives showed strong enhancement of activity against
glycosidases as compared to parent monovalent derivatives.
For instance, Fig. 5 depicts fullerene-based dodecavalent derivative 40,
showing roughly 2150 times improvement with respect to monovalent imino
sugar 39 against a-mannosidase.
222 | Carbohydr. Chem., 2012, 38, 215–262

Boc
N O
HO
1) NH3 (aq.) NH
2) (COCl2)2
Cl
3) TFA
O
OMe 4) CuCl HO NH2
TBDPSO
O 5) HCl
Me
HO
Me
OMe Isofagomidine hydrochloride
41 42 (47%, 5 steps) Ki 0.75 μM (α-Mannosidase
1) TBAF from jack beans)
2) [O]
+
3) H H
N O
O
OH
HO
OH
43 (79%)
Ki 36 nM (bovine liver β-glucuronidase)
Scheme 5
3.2 Six-membered amidines

Although the use of amidines, amidoximes, and amidrazones as glycosidase
inhibitors was developed by Ganem in the 1990’s,1 no previous reports
concerning the use of such functional groups at the pseudoanomeric posi-
tion are found in the literature. In this context, Bols and co-workers
reported36 on the synthesis of isofagomidine 42 (Scheme 5, 16 steps from
D-arabinose). The key steps are the introduction of an aminomethyl group
at C-4 of D-arabinose, the conversion of C-1 into a nitrile moiety and
copper-catalyzed cyclization of an amino group on the nitrile, giving the
amidino scaffold.
Surprisingly, isofagomidine hydrochloride behaved as a strong inhibitor
of a-mannosidase (Ki=0.75 mM), while being a poor inhibitor of b-gluco-
sidase, regardless the pH of the assay and the incubation time. These results
strongly contrast with the inhibitory profile of related isofagomine and
isofagomine lactam, low-micromolar inhibitors of b-glucosidase, despite
having the three compounds a related structure. Calculated atomic and
group charges on isofagomidine, isofagomine and its lactam indicated
electronic differences that might support the unexpected different inhibitory
profile.
Protected lactam 41 (prepared in 11 steps from D-arabinose) was also
efficiently transformed37 in three steps (silyl group removal, oxidation of the
primary alcohol, deprotection) into 43 (Scheme 5). Compound 43 is the
uronic acid analogue of isofagomine lactam, and turned out to be a
nanomolar glucuronidase inhibitor (Ki=36 nM).
3.3 Azafagomines
Azafagomine 51 (Table 1) is a polyhydroxylated cyclic hydrazine with the
D-gluco configuration having a hybrid structure between an imino sugar and
Carbohydr. Chem., 2012, 38, 215–262 | 223

an azasugar.1 In the search for more potent and selective inhibitors
structurally related to azafagomine, López and Bols considered38 the
possibility of a selective alkylation of 1-azafagomine at the 2-N position; the
N-1 was previously alkylated by Bols’ group leading to compounds with
reduced activity.39 Furthermore, alkylation of the pseudoendocyclic posi-
tion of isofagomine (equivalent to N-2 in azafagomine) was reported to
furnish b-glucocerebrosidase inhibitors in the low nanomolar range.40
Selective N2-alkylation of azafagomine was conducted by reductive
amination involving a series of aldehydes and the free nitrogen of partially
protected azafagomine 44 (Scheme 6).41
Fully unprotected derivatives 45–50 were tested as inhibitors of yeast
a-glucosidase and b-glucosidase from almonds; inhibitory tests were con-
ducted at pH 6.8 at which all of them should be largely unprotonated.42
Inhibitory data (selected values are presented in Table 1) indicated that
substitution at the 2-N position with fatty chains strongly decreased activity
against a-glucosidase, suggesting that this kind of substitution disrupts
interaction with the active site. Remarkably, a strong improvement in the
inhibition of b-glucosidase, especially for the N-phenylpropyl derivative 50
(Ki=32 nM), was achieved. These results are probably due to stabilizing
interactions with hydrophobic residues located on the active site of b-
glucosidase. Therefore the overall effect is a strong selectivity increase
towards the latter enzyme.
Alves et al. have accomplished43 a novel preparation of both enantiomers
of azafagomine, together with their 2-N-phenyl carboxamide derivatives.
This synthesis was based on the original preparation of enantiopure
1-azafagomine reported by Bols,44 in which the key step consisted of a
hetero-Diels-Alder reaction between achiral penta-2,4-dien-1-ol and
4-methyl-1,2,4-triazoline-3,5-dione, followed by kinetic resolution of both
enantiomeric cycloadducts with lipases. Alves43 replaced the achiral
pentadienol with 2,4-pentanedionates bearing tetra-O-acetyl glucosyl chiral
auxiliary, which allowed an unprecedented enantiomeric synthesis of
azafagomine and its derivatives (Scheme 7).
OBn OH
1) R-CHO
8 steps NH H2, Pd(OH)2 N
BnO HO R
L-Xylose BnO 2) aq. HCl, Δ HO
NAc 3) aq. NH NH
3
44 45–50
R = Butyl, hexyl, heptyl, nonyl, decyl,

3-phenylpropyl
Scheme 6
EtO2C
OAc O
O
O OH Ph
AcO + N
AcO O 1) Et3SiH, TFA N 1) H2O, H N
HO H
O N Ph
HO
OAc CO2Et 2) Oxone N 2) LiAlH4 NH
N N
40%, 2 steps 15%, 2 steps
52 O 53 O 54
N O
Ph
Scheme 7
224 | Carbohydr. Chem., 2012, 38, 215–262

Table 1 Inhibition profile for azafagomine 51, N-alkylated azafagomines 46, 48, 50 and
N-acylated azafagomine 54 (Ki in mM, pH 6.8).
OH
N
HO R
HO
NH
Compound R a-Glucosidase b-Glucosidase a/b-Selectvity
51 H 6.90 0.32 22
46 278 0.055 5054

5
48 169 0.140 1141

8
50 Ph 158 0.032 4938

O
54 Ph 3.36a 14.7a 0.23
N
H
a
pH 7.0.
·HCl NH NH NH O
N N N
H CBz H
OH OH N
H
HO O HO
OH O OH
56 55 57
Ki 4.2 μM (α-L-fucosidase, Ki 1.0 μM (α-L-fucosidase,
bovine kidney) bovine kidney)
D-Lyxose
Scheme 8
Removal of the chiral auxiliary from cycloadduct 52 (obtained from

4-phenyl-1,2,4-triazole-3,5-dione and the corresponding 2,4-pentadienoate
bearing a tetra-O-acetyl-b-D-glucopyranosyl moiety), followed by diaster-
eoselective epoxidation, opening of the oxirane and reductive cleavage with
LiAlH4 afforded phenyl carboxamide 54 (Scheme 7).
Inhibition assays of 54 revealed43 (Table 1) that, unlike the 2-N-alkylated
analogues 45–50, acylation of the same position afforded an improvement
of the inhibition against a-glucosidase (Ki=3.36 mM), and an impairment
against b-glucosidase as compared with parent azafagomine 51. Binding of
54 with the active site of yeast a-glucosidase was studied by molecular
modelling, suggesting an efficient packing of the carboxamido moiety on a
hydrophobic subsite of the enzyme.
Moreno-Clavijo et al. have recently reported45 the first examples of
L-fuco-configured azafagomines bearing substituents at the C-3, considered
as diaza-C-glycosyl analogues. A library of such compounds was accessed
starting from D-lyxose (Scheme 8), using the reductive hydrazination of a
1-deoxy-ketohexose with tert-butyl carbazate as the key step, in a similar
Carbohydr. Chem., 2012, 38, 215–262 | 225

way as in the enantioselective preparation of azafagomine by Bols via
partially-protected reducing L-xylofuranose.41 Compounds 56 and 57,
bearing either a hydroxymethyl moiety or a biphenyl-based amide linkage,
exhibited potent and selective inhibition against a-L-fucosidase, within the
same order of magnitude as parent unsubstituted L-fuco-azafagomine.46
4 Azepanes and azetidines

Although most imino sugars found in the literature are five- or six-mem-
bered ring compounds, some other different ring sizes have been designed
and tested as potential glycosidase inhibitors.
Among them, unnatural polyhydroxyazepanes (seven-membered imino
sugars) are the most remarkable examples; since pioneering work of Wong47
in the mid-1990’s, a great number of polyhydroxylated azepanes have been
reported, including bicyclic derivatives.
One of the most relevant properties concerning azepanes as glycosidase
inhibitors is their higher flexibility as compared to pyrrolidines and piper-
idines; azepanes can adopt more easily a quasi-flattened conformation, and
thus might better interact with the active site of glycosidases.47 Due to the
fact that the synthesis of these compounds has been less developed than
their five- and six-membered counterparts, the molecular requirements for
predicting good inhibitory activities of azepanes still remain unknown.
In this context, Blériot’s group has accomplished the synthesis and eva-
luation of a large number of azepanes, by modifying the number of the
hydroxyl groups, together with the configurations of the sugar residue, and
the substituents on the ring.
For example, benzylated methyl 6-azido-6-deoxyglucopyranoside 58 was
converted,48 via a tandem Staudinger-aza-Wittig approach, (Scheme 9) into
partially protected azepane 60. In this approach, the key step is the treat-
ment of azido lactol 59 with triphenylphosphane, followed by intramole-
cular reaction of the transient iminophosphorane with the latent aldehyde
to give an imine which upon reduction with NaBH3CN furnished the aze-
pane architecture. b-Hydroxyazepane 60 was an efficient building block in
the preparation of three different families of seven-membered ring imino
sugars bearing either hydroxymethyl, acetamido or carboxy moieties loca-
ted at the b position to the endocyclic nitrogen (Scheme 9); such compounds
are considered as ring homologues of 1-N-imino sugars.
The hydroxymethyl moiety present in derivatives 64 and 65 was intro-
duced by oxidation of the free hydroxyl group, followed by Wittig olefi-
nation and hydroboration-oxidation process (Scheme 9). Deprotection of
C-6 epimers 62 and 63 furnished compounds 64 and 65.
Azepanes 64 and 65 were found to be rather potent and selective
uncompetitive inhibitors (Table 2) of b-glucosidase (Ki=4 mM) and a-
mannosidase (Ki=42 mM), respectively.
Blériot’s group also attempted the synthesis of stable D-manno-configured
polyhydroxylated azepanes, which may act as analogues of noeuromycin
138 (Fig. 10). For that purpose, the C-2 epimer of azidolactol 59 was
used;49 in this case, trihydroxylated D-manno and L-gulo derivatives, bearing
a methyl group at the C-6 position were obtained and found to be
226 | Carbohydr. Chem., 2012, 38, 215–262

N3 N3 CBz H
1) Ac2O, N N
O H2SO4 O 1) MsCl, Py
BnO BnO
BnO BnO
2) NaN3, Δ
2) NaOMe OH OBn OH
OBn HO 3) PPh3, H2O AcHN
OBn
OMe 4) Ac2O, Py
BnO OBn HO OH
58 59 5) H2, Pd/C, HCl
60 67
1) PCC
CBz 2) Wittig olefination CBz CBz
N N N
HO HO
1) BH3
63 OBn OBn OBn
2) H2O2, NaOH +
1) RuCl3, NaIO4 BnO OBn BnO OBn BnO OBn
2) H2, Pd/C, HCl
61 62 63
H·HCl 3 : 2
N
H2, Pd/C, HCl H2, Pd/C, HCl
HOOC OH
H H
HO OH N N
66
HO HO
43%, overall yield OH OH
+
HO OH HO OH
64 65
Scheme 9
Carbohydr. Chem., 2012, 38, 215–262 | 227

potent although non-selective inhibitors of a-glucosidase, b-glucosidase
and a-L-fucosidase, with a lower potency than the isomers depicted in
Scheme 9.
Treatment of derivative 63 with RuCl3 and NaIO4, followed by depro-
tection, furnished derivative 66, which incorporates a carboxylic acid
moiety; this compound was unexpectedly found to be the most potent
azepane-based inhibitor of a-L-fucosidase (41 nM, Table 2),50 despite
lacking both, the CH3 group and the L-fuco configuration. These results
clearly indicate how the conformational flexibility of seven-membered ring-
based imino sugars might allow strong interactions with enzymes despite
even lacking the appropriate configuration on the sugar residue.
Compound 60 was also used as the key intermediate for the preparation
of azepanes bearing acetamido groups (Scheme 9); the free hydroxyl group
of 60 was transformed into the acetamido moiety upon mesylation,
nucleophilic displacement of the mesyloxy group by sodium azide, Stau-
dinger reduction and acetylation of the corresponding amine to give 67.51
This compound, with the D-gluco configuration, is considered as a ring
homologue of 2-acetamido-1-N-imino sugars, being a potent and selective
inhibitor of b-N-acetylglucosaminidases, with Ki values in the sub-
micromolar range (0.4 and 0.7 mM). Moreover, compound 67 was found52
to be a strong competitive inhibitor of human O-GlcNAse (Ki=11 mM) and
BtGH84 (Ki=8mM). Crystal structure of 67 bound to BtGH84 revealed a
slightly distorted 2B5,6 conformation, similar to rare 4E and 1,4B con-
formations adopted by complexes of GlcNAc-enzymes on mutant chit-
inases53 and b-N-acetylglucosaminidases,54 and quite different to those
conformations found in solution (4,5CN,2 and 5,6T2,3 in an almost 1:1
ratio).52
Blériot’s group also reported51 the preparation of polyhydroxylated
azepanes 70 and 71, compounds bearing both, an acetamido group at the
g- positions to the endocyclic nitrogen, and a hydroxymethyl moiety. Title
compounds were obtained from D-arabinose, using azacycloheptene 68 as
CBz CBz
N N
BnO BnO
D-Arabinose
BnO X
BnO
BnO BnO Y
X = OH, N3
68 1) PPh3, H2O 69 Y = N3, OH
2) Ac2O
3) H2, Pd/C, HCl
H H
N N
HO HO
OH HO OH
HO
HO NHAc HO NHAc
70 71
Scheme 10
228 | Carbohydr. Chem., 2012, 38, 215–262

Table 2 Inhibition profile for azepanes 64–67, 70, 71 (Ki in mM or inhibition %).
b-N- b-N-
a-L-fucosidase Acetilglucos- Acetilgluco-
b-Glucosidase a-Mannosidase (bovine aminidase saminidase
Compound (almonds) (jack bean) epididymis) (jack bean) (bovine liver)
64 4 33% N.I. (from N.I. N.I.

bovine
kidney)
65 Not tested 42 Not tested Not tested Not tested
66 N.I. N.I. 0.041 Not tested Not tested
67 Not tested N.I. 28% 0.4 0.7
70 Not tested 25% N.I. 0.5 0.6
71 Not tested N.I. N.I. 0.05 0.4
N.I. – no inhibition.
the key intermediate; such derivative was transformed into 69 by epoxida-

tion, followed by azide opening of the oxirane ring (Scheme 10). Access to
70 and 71 was accomplished by Staudinger reduction of the azido group,
followed by chemoselective acetylation and deprotection. Azepanes 70 and
71 turned out to be potent hexosaminidase inhibitors (Table 2).
In this case, unlike other related azepanes, there was a clear correlation
between the relative configuration of the imino sugar and the target enzyme,
as b-D-gluco-configured 70 was a much stronger inhibitor against b-N-acetyl
glucosaminidases than a-L-ido and a-D-manno counterparts. Nevertheless,
this observation was not fulfilled for b-L-gulo azepane 71, which turned
out to be an inhibitor ten times stronger than 70 against the same enzyme
(Table 2).
Dhavale’s group recently reported55 the efficient preparation of tri- and
tetrahydroxyazepanes starting from inexpensive and readily available
D-glucurono-6,3-lactone (73). In particular, Scheme 11 depicts the synthetic
pathway for preparing tetrahydroxy derivative 76 upon DIBAL reduction
of the lactone moiety, reductive amination, N-protection, intramolecular
cyclization, and deprotection. Azepane 76 was screened against 7 glycosi-
dases, turning out to be a potent and selective a-galactosidase inhibitor
(Ki=5 mM).
56
D-Quinic acid was used by Shih et al. as starting material in another
procedure for preparing azepanes; Scheme 12 depicts the preparation of
trihydroxylated azepane 82, bearing a n-butyl residue at the C-7 (7R con-
figuration).57 Enone 77, obtained in four steps from D-quinic acid was
subjected to 1,2-addition with butyl lithium promoted by CeCl3, giving
compound 78 as the major diastereoisomer.
Derivative 78 underwent allylic oxidation with PCC, followed by
Luche reduction to afford allylic alcohol 79 with complete stereoselectivity.
Protection of the free hydroxyl group, cis-dihydroxylation of the double
bond, oxidative cleavage and aminocyclization furnished the azepane
architecture. Separation of epimeric 80 and 81, followed by deprotec-
tion gave 82, which turned out to be a potent b-galactosidase inhibitor
(Ki=3 mM).
Carbohydr. Chem., 2012, 38, 215–262 | 229

O O HO
OO
HO OO OO
OH HO HO
H2SO4
DIBAL-H
Acetone
O O
95%
OH O O
72 73 74
CBz
1) BnNH2, H+
NBn
·HCl NaBH3CN
N 1) TFA (aq.) HO O
2) CBzCl, base
2) H2, Pd/OH)2 OH
HO OH 3) HCl
O
HO OH 84% (3 steps) O
88% (3 steps)
76 75
Ki 5 μM (α-galactosidase
from green coffee beans)
Scheme 11
O Bu OH
Bu
4 steps BuLi O 1) PCC

O
D-Quinic acid OH
O CeCl3 O O
2) NaBH4, CeCl3·7H2O
O
70%
41% overall yield
77 78 79
1) MOMCl
H 2) KMnO4
Bu N 3) NaIO4
H2, Pd/C O O 4) BnNH2, NaBH3CN
HCl
OH
MOMO
R2
HO OH N
Bn R1
82 (90%)
IC50 = 3μM (β-Galactosidase) 80 R1 = Bu, R2 = H (22%)
81 R1 = H, R2 = Bu (12%)
Scheme 12
Up to date, reports concerning the synthesis of hydroxylated azetidines

are scarce; the first search for biological activity of these carbohydrate
mimics was reported by Jäger58 in the late 1990’s. More recently, Fleet’s
group has developed59 an efficient procedure for preparing azetidine-based
imino sugars which behaved from moderate to good inhibitors of several
glycosidases. The key building blocks are stable 3,5-di-O-triflates derived
from pentoses, which undergo a double nucleophilic substitution reaction
with amines to afford the azetidine ring. For example, azetidine 87
was found to be a good inhibitor of non-mammalian amyloglucosidase,
a-glucosidase and b-galactosidase (Scheme 13).
230 | Carbohydr. Chem., 2012, 38, 215–262

O O O
HO O O O
TfO R-NH2
Tf2O, Py
–30 ºC N
HO O TfO O O
93%
83 84 Bn
85 (87%)
R = Bn
OH
OH HO 1) aq. TFA
HO NH4HCO2
HO 2) NaBH4
HO Pd/C
82%
97% N
N
H
Bn
87 86
IC50 (μM) 19 (Amyloglucosidase from Rizopus sp.)

39 (α-Glucosidase form A. niger)
70 (β-Galactosidase from rat intestinal lactase)
Scheme 13
5 Bicyclic imino sugars

5.1 Indolizidine, pyrrolizidine, quinolizidine and nortropane analogues
Up to date, many bicyclic imino sugars have been isolated from nature60
and tested as glycosidase inhibitors; some of them with a strong activity, in
many occasions against a-glucosidases. It has been claimed that probably
their intrinsic rigid structure is responsible for the biological activity exerted
by many of these carbohydrate mimics.61,62
Compounds like lentiginosine, swainsonine, and castanospermine (di, tri
and tetrahydroxylated indolizidines), casuarine, and hyacinthacines (poly-
hydroxylated pyrrolizidines), and quinolizidines are remarkable examples.1
A novel indolizidine alkaloid, ()-steviamine 88, has recently been isolated
from the leaf material of Stevia rebaudiana.63 Its enantiomer 89 and 5-epi-
steviamine 90 (Fig. 6) have been synthesized from a D-ribose-derived cyclic
nitrone.64 ()-Steviamine showed weak inhibition of b-glucosidases, but
good activity against b-galactosidase (IC50=35 mM), and no inhibition of
either a-mannosidase or a-galactosidase, despite its structural resemblance
with swainsonine. Interestingly, 88 also inhibited a-galactosaminidase
(GalNAcase); although the activity for this latter enzyme was found to be
moderate, it was the first case of a natural product inhibiting a GalNAcase.64
Despite the potency exhibited by some of the natural imino sugars like
deoxynojirimycin and castanospermine, they frequently lack selectivity. For
instance, these compounds are potent inhibitors of a-glucosidases at the
endoplasmic reticulum, but they also bind lysosomal a- and b-glucosidases
and the intestinal glucoamylase, a real drawback in their potential appli-
cation as therapeutic agents.
The increase of the sp2-character of the endocyclic nitrogen atom has
become an alternative to improve the anomeric selectivity of the corre-
sponding sugar mimic. In this context, Cipolla et al. reported62 the synthesis
of novel nojirimycin derivatives containing cyclic carbamates, ureas, and
guanidines; nevertheless, the observed inhibitory activities were moderate.
Thiohydantoin-castanospermine analogues, where the pyrrolidine ring of
castanospermine is replaced by a thiohydantoin scaffold, were recently pre-
pared as a new family of sp2-imino sugar-type glycomimetics.65 It was
Carbohydr. Chem., 2012, 38, 215–262 | 231

H 3C H 3C H 3C
HO HO HO
N N N
HO OH HO OH HO OH
88 (–)-Steviamine 89 (+)-Steviamine 90 5-epi-(+)-Steviamine

IC50 35 μM (rat β-galactosidase) IC50 484 μM (α-rhamnosidase) IC50 342 μM (α-rhamnosidase)
Fig. 6
OH
MeOOC
H H
N i) TEMPO, NaOCl N
ii) DCBzCl, MeOH
O O
N3 N3
O O
PMBO PMBO
91 92
1) H2, Pd/C
2) Octyl-NCS
3) DDQ O
HO
N (CH2)7 CH3
N
HO
S
OH
93
Ki 59 μM
(β-glucosidase from almonds)
Scheme 14
claimed that delocalization of the nitrogen electron pair generates a partial

positive charge on the nitrogen that might improve mimicry of the transition
state of the enzyme-mediated glycoside hydrolysis.66 Scheme 14 depicts the
synthetic pathway for accessing 93, an example of a thiohydantoin-casta-
nospermine hybrid structure. Precursor 91 was converted into the a-azido
ester 92, followed by subsequent azido reduction, coupling with an alkyl
(e.g. octyl) isothiocyanate and deprotection to afford glycomimetic 93. Title
compound was found to be a selective and good inhibitor of b-glucosidase.
In a similar fashion, other sp2-imino sugars resembling castanospermine
were prepared; bicyclic iso(thio)urea castanospermine analogues 94–97
(Fig. 7) represent good examples of this approach.67 Such compounds were
accessed starting from 3-O-acetyl-5-azido-5-deoxy-1,2-O-isopropylidene-a-
D-galactofuranose; the key intermediate was a carbodiimide, obtained by
treatment of the parent thiourea with HgO.
Regarding compounds 94–97, the shift of their inhibitory activities from
a- to b-glucosidase is remarkable, when compared to castanospermine; such
data are even more surprising considering that, due to a strong anomeric
effect, derivatives 94–97 exist only as the a-anomers.
232 | Carbohydr. Chem., 2012, 38, 215–262

Z N
Y 94: X = OH, Y = H, Z = O; Ki = 1.90 μM

N 95: X = H, Y = OH, Z = O; Ki = 0.019 μM
X 96: X = OH, Y = H, Z = S; Ki = 0.76 μM
HO
97: X = H, Y = OH, Z = S; Ki = 0.023 μM
OH 98: X = OH, Y = H, Z = NH; Ki = 0.42 μM
OH (β−Glucosidase from almonds)
Fig. 7
Furthermore, both, iminooxazolidines 94, 95 and iminothiazolidines

96, 97, with pKa values of virtually 7.0, were found to be more potent
inhibitors at neutral pH than at acidic pH; this fact was explained con-
sidering a better interaction of partially protonated imino sugar (roughly
50% at pH 7.3) with fully deprotonated catalytic system at the active site of
the enzyme.67 Thermodynamic studies revealed an entropically-driving
binding process with b-glucosidase from Thermotoga maritime (TmGH1),
quite an uncommon situation.
Additionally, Brumshtein et al.68 accomplished the synthesis of 98
(Fig. 7); this compound incorporated a strongly basic residue (guanidino
group) on the castanospermine-based template, in order to study potential
favorable coulombic interactions with the active site of glycosidases.
The synthesis started from 5-azido-5-deoxy-1,2-O-isopropylidene-a-D-
glucofuranose (99) and was based on the introduction of a second azido
group at the C-6; reduction of both azido moieties yielded diamine 100,
which was transformed into imidazolidine-2-thione 101 by thiocarbonyla-
tion with CS2 and then into the S-methyl isothiouronium salt 102;
subsequent nucleophilic displacement of the methylthio group with n-
octylamine and deprotection gave target compound 98 (Scheme 15).
Like derivatives 94–97, compound 98 also showed a high selectivity for
b-glucosidase.
Another approach investigated by the same group to improve the selec-
tivity and permeability of imino sugars was the incorporation of axially-
anchored alkyl chains at the pseudoanomeric position of carbamate-derived
castanospermines (Fig. 8). With that purpose, 5-N-octyl (104), 5-S-octyl
(105) and 5-C-octyl (106) derivatives of sp2-imino sugar-type castanos-
permine 103 have been prepared and tested.69
Title compounds were found to be selective and potent inhibitors of
neutral a-glucosidase, and remarkably, none of them was active against the
human lysosomal acid a-glucosidase. Encouraging preliminary results on
breast cancer cell (MCF-7 cell line) suggested their potential use as anti-
proliferative agents.
Concerning polyhydroxylated quinolizidines, Pandey et al. described70 a
divergent route starting from D-ribose (Scheme 16 displays the retro-
synthetic analysis) to get a small library of quinolizidine alkaloids (111–
123). Among them, two compounds had remarkable inhibitory properties:
112, against a-galactosidase from coffee beans (Ki=83.9 mM), and 123,
against a-glucosidase from yeast (Ki=28 mM).
Among polyhydroxylated pyrrolizidines, casuarine derivatives have
become the most investigated structures in this area in the last few years.
Carbohydr. Chem., 2012, 38, 215–262 | 233

OH NH2
O 1) TsCl, py O
2) NaN3 O CS2
O
3) H2, Pd/C DCC
N3 H 2N
O O
HO HO
99 100
S S I
NH NH
O O
HN HN
O MeI O
H H
O O
HO HO
101 102
1) n-Octyl-NH2
98 2) TFA
Ki (μM)
27 (β-glucosidase from almond, pH 5.5)
0.42 (β-glucosidase from almond, pH 7.3)
35 (β-glucosidase/galactosidase from bovine liver)
0.18 (naringinase from Penicillium decumbens)
Scheme 15
O O O
O
104 X=NH n = 6 Ki 0.54 μM
N 105 X=S n = 6 Ki 3.4 μM
HO N
HO 106 X=CH2 n = 5 Ki 0.87 μM
HO HO (Neutral α-glucosidase II ER, yeast)
OH OH
103 OH X
n
Fig. 8
Casuarine 128 and its 6-O-a-glucoside 129 were isolated from the bark of
Casuarina equisetifolia. Both compounds are potent inhibitors of fungal
glucoamylase from Aspergillus niger (IC50=0.7 and 1.1 mM, respectively);
furthermore, they were recently found to act as strong and selective inhi-
bitors of human N-terminal subunit maltase-glucoamylase (NtMGAM),
with Ki values of 0.45 and 280 mM respectively.71 This enzyme is involved in
starch digestion, so inhibitors may potentially be used in the treatment of
diabetes.
Total syntheses of 128, 12971 (Scheme 17) and other casuarine analogues
(130 and 132), together with their 6-O-a-glucosides (131 and 133)72 have
been reported; such syntheses involved stereoselective 1,3-dipolar cycload-
ditions of a chiral cyclic nitrone and an alkene, Tamao-Fleming oxidation
and regioselective glucosylation. Their biological evaluation demonstrated
that casuarine derivatives bind tightly and selectively to glucoamylase from
Aspergillus niger.
234 | Carbohydr. Chem., 2012, 38, 215–262

R1 R R1 R
H H
HO HO
N N
HO HO
R1 R
D-gluco type azasugars L-gluco type azasugars
111 OH CH2OH 118
112 H OH 119
113 OH CH3 120
114 NH2 H 121
115 NHC14H29 H 122
H H
116 NHBn H 123
O 117 NHC12H25 H O
N N
O O
107 109
O O
D-Ribose
N N
O O
108 110
Scheme 16
Casuarine 128 and specially glucoside 129 were also found to be very
strong trehalase inhibitors;71 trehalase is the enzyme that catalyzes breaking
down of trehalose, the vital carbohydrate source for insects. Therefore,
trehalase inhibitors are claimed to be potential insecticidal agents.73
Natural (þ)-hyacinthacine A1 134 (Fig. 9) was isolated from Muscari
ormeniacum bulbs.74 Since then, several syntheses of this compound have
been reported75 due to its good inhibitory properties against b-galactosidase
(IC50=4.4 mM), a-L-fucosidase (IC50=46 mM) and amyloglucosidase
(IC50=2.3 mM).74 D’Adamio et al.76 achieved the total synthesis of 134,
(þ)-7a-epi-hyacinthacine A1 135 (Fig. 9), and their 6-hydroxy analogues,
like 136. The synthesis involved a nitrone cycloaddition strategy, with a
D-ribose-derived cyclic nitrone as the dipole, and tert-butyl acrylate as the
dipolarophile.
Another type of relevant bicyclic imino sugars are calystegines; these
natural polyhydroxy-nor-tropane alkaloids (named after Calystegia sepium)
combine both, a pyperidine and pyrrolidine ring in their structure, and are
potent and quite selective b-glucosidase inhibitors.77 This fairly good inhi-
bitory profile results from good mimicry of the transition state of the sub-
strate hydrolysis. It has been observed that, upon interaction with
glucosidases, calystegines adopt a conformation that places the endocyclic
nitrogen in the same position as the O-5 of the glycosides, the natural
enzymatic substrates.78,79
Further modifications of natural calystegines may provide interesting
inhibitors. In this context, Rasmussen and Jensen envisioned noeurostegine
139 as a new stable hemiaminal with a nor-tropane motif,77 and a hybrid
structure between noeuromycin 13880 and calystegine B2 137 (Fig. 10).
Carbohydr. Chem., 2012, 38, 215–262 | 235

CO2Et OH
O–
HO
SiMe2Ph O SiMe2Ph
N+ CO2Et O
Zn, AcOH/H2O
+ N H N H
60–65 ºC
SiMe2Ph
BnO OBn OBn OBn
124 126
OBn OBn
OBn OBn
OH
125
O
HO OH OH
HO
OH
O OBn 6 OH
129 Casuarine O 5 7
6-O-α-glucoside
B
Ki 12 nM (Tre37A) N H N H
Ki 0.66 nM (C. riparius) OH
3 A 1
OBn 2 OH
N H
OBn OH
OBn OH
OH
127 128 Casuarine
Ki 0.45 μM (NtMGAM)
OH
OH
OH HOH2C OH
H H
RO OH RO OH
N N
OH OH
130 R = H Ki 3.5 μM (glucoamylase) 132 R = H Ki 11 μM (glucoamylase)

131 R = α-D-Glc Ki 7.4 μM (glucoamylase) 133 R = α-D-Glc Ki 23 μM (glucoamylase)
Ki 86 nM (Tre37A) Ki 2.8 μM (Tre37A)
Ki 22 nM (C. riparius) Ki 157 nM (C. riparius)
Scheme 17
OH OH OH
H H H
OH OH HO OH
N N N
OH OH OH
134 135 136

IC50 4.4 μM (β-galactosidase) IC50 346 μM IC50 7.7 μM
IC50 46 μM (α-L-fucosidase) (Amyloglucosidase) (Amyloglucosidase)
IC50 2.3 μM (amyloglucosidase)
Fig. 9
236 | Carbohydr. Chem., 2012, 38, 215–262

Noeuromycin is a strong glycosidase inhibitor, but it incorporates an unstable
hemi-aminal functionality. Thus, noeurostegine actually combine the stability
of calystegines and the potency of noeuromycin against b-glucosidase, while
exhibiting a much higher selectivity; derivative 139 is virtually non-active
against a-glucosidase, and much less active against other glycosidases.77
The synthesis of neurostegine is depicted in Scheme 18; compound 139
was obtained in 22 steps starting from levoglucosan; one of the key steps
was the transformation of nonadiene 143 into cycloheptene 144 via a ring
closing metathesis reaction.
Alternatively, compounds with the structure of 6-oxa-calystegine, where
the nitrogen atom has a sp2-character, were postulated as a good approach
to get potent and selective glycosidase inhibitors, in a similar fashion as
reported for castanospermine.
For that purpose, thioureas 140 and 141 (6-oxa-N-thiocarbamoyl-calys-
tegines, Fig. 10),81 were prepared from 5-amino-5-deoxy-1,2-O-iso-
propylidene-b-L-idofuranose 146 (Scheme 19). The synthetic pathway
involved coupling of 146 with commercially-available n-butyl and n-octyl
isothiocyanates, followed by deprotection of the intermediate thioureas 147
and 148. 6-Oxa-calystegines 140 and 141 were found to be potent and
selective inhibitors of b-glucosidase (Thermotoga maritima), with Ki values of
1.1 mM and 0.28 mM, respectively. n-Octyl derivative was also quite active
against human b-glucocerebrosidase (Ki=2.2 mM), a tenfold better than 140,
demonstrating again that increased lipophilicity considerably improves the
potency and selectivity of certain imino sugars. Regarding compound 141,
with a more lipophilic structure, thermodynamic data indicated favorable
enthalpic and entropic contributions to binding, probably because such
compound displaces more water molecules from the enzyme active site.81
Chagnault and co-workers accomplished82 preparation of 10-azabi-
cyclo[4.3.1]decanes, a different family of bicyclic imino sugars. The best
inhibitory profile was observed for the pentahydroxylated nortropane 149
(Fig. 11) against b-glucocerebrosidase (IC50=1.2 mM) and b-glucosidase
(IC50=2.0 mM).
HO OH
HO HO
HO NH HO NH
OH OH
Calystegine B2 137 Noeuromycin 138

Ki 1200 nM Ki 69 nM (β-glucosidase from almonds)
(β-glucosidase from almonds) Ki 20 nM (α-glucosidase from yeast)
O
OH HO S
HO
HO HO N
HO NH N
OH n
OH H
Noeurostegine 139 140 n = 3 Ki 1.1 μM, (β-glucosidase from
Ki 50 nM (β-glucosidase from almonds) Thermotoga maritima)
141 n = 7 Ki 0.28 μM, (β-glucosidase from
Thermotoga maritima), Ki 2.2 μM, (Gcase)
Fig. 10
Carbohydr. Chem., 2012, 38, 215–262 | 237

O O
BnO
OH 8 steps OBn 5 steps
O O
BnO
OH OH OBn OPMB
OBn
142 143
Levoglucosan
OBn
Grubbs-Hoveyda
BzO catalyst
OH
BnO BnO
4 steps 4 steps
HO
HO
NH
BnO BnO
OH
Noeurostegine 139 N3 OPMB
145 144
OBn OBn
Scheme 18
OH OH
S
O O
O O
H 2N R-NCS R N
N H
H
O O
HO HO
146 FA 147 R = butyl
.T
aq 148 R = octyl
140 R = butyl
141 R = octyl
Scheme 19
HO
OH
HO NH
OH
HO
149
IC50 1.2 μM (β-glucocerebrosidase)
IC50 2.0 μM (β-glucosidase)
Fig. 11
5.2 Thiazolines
In the mid-1990’s, Knapp’s and Vocadlo’s groups reported the first exam-
ples of bicyclic thiazolines as good inhibitors of hexosaminidases, a family
of enzymes that exert their catalytic activity trough oxazolinium ion 151
(Fig. 12). Bicyclic thiazoline 153 (GlcNAc thiazoline) was accessed by
238 | Carbohydr. Chem., 2012, 38, 215–262

R O O
HO O HO OH
HO H R
O H2 O HO
HO O O
HO HO
HO O
O HO HN O
HN
O
N+
H CH3
CH3
CH3
150 151 152
OH OH
OH OH
O O S
HO
HO O CH3
HO HO
S N N
S
N OH
CH3 CH3
0
S2 4
C1
1
S3
153
GlcNAc thiazoline.
Ki 20 μM (SpHex)
Ki 70 nM (human Hex)
Ki 70 nm (O-GlcNAcase)
Fig. 12
intramolecular cyclization involving the anomeric carbon atom and a

thioamide scaffold located at the C-2 of per-O-acylated N-acetil glucosa-
mine.83 This compound is a potent inhibitor of retaining N-acetylhex-
osaminidases, for which it is considered as a good transition state
analogue;84 recent conformational studies conducted on GlcNAc thiazoline
suggested that 153 binds the active site of the enzyme in a distorted 4C1 chair
conformation, the conformation that most closely resembles transition state
151. NMR data and theoretical calculations demonstrated that, in solution,
compound 153 can move among one chair conformation (4C1) and two skew
boat conformations (1S3 and 0S2, the latter being the most populated one).85
Modification of 153, involving imine-to-enamine tautomeric processes,
followed by electrophilic or radical additions (e.g. of thiols) has been
reported (Scheme 20). Using this methodology, compound 159 was pre-
pared and turned out to possess an excellent selectivity for human
O-GlcNAcases, and to be almost as potent as parent 153.86 Moreover,
chitobiose and chitotriose thiazolines, in which the corresponding oligo-
saccharide bears a GlcNAc thiazoline residue at the terminal unit, have
recently been reported as excellent chitinase inhibitors of Chi18A from
Serratia marcescens (Ki=25 and 0.25 mM respectively).87
Other derivatives of GlcNAc and GalNAc thiazolines are the C-1
homologated analogues 160–162 (Fig. 13). The inhibition of these com-
pounds against N-acetyl-b-hexosaminidases from Streptomyces plicatus
(SpHex), falls dramatically when increasing the size of the C-1 substituent.
This observation suggests strong steric restrictions around the anomeric
position for binding the enzyme.88
Carbohydr. Chem., 2012, 38, 215–262 | 239

OAc OAc OAc
O TfOH O
O AcO AcO
AcO P4S10 pyridine
OAc AcO
AcO AcO
NHAc N S HN S
CH3
154 155 156
I2, THF
OH OAc OAc
O O NaN3, O
HO deacetylation AcO AcO
DMF
HO AcO AcO
N S N S N S
N3 N3 I
159 158 157
Scheme 20
OH OH OH OH
O O O
HO HO
CH3 HO
HO HO NHR
N N N Se
S S
CH3 CH3 CH3

160 161 R=H Ki >2 mM (SpHex) 163
Ki 200 μM (SpHex) 162 R=Boc Ki >2 mM (SpHex) GlcNAc selenazoline
Ki 170 μM (SpHex)
Ki 0.7 μM (GlcNAcase)
Fig. 13
Replacement of the sulfur atom of thiazoline 153 with selenium led to

GlcNAc selenazoline 163, showing much weaker inhibition of hex-
osaminidase than 153 (Figs. 12 and 13 for data).89
5.3 Carbohydrate-derived inhibitors bearing spiranic motifs

It has been demonstrated that in many occasions there is a close correlation
between reduced conformational flexibility and a potent interaction with
biomolecules.90 In this context, natural hydantocidin 164 (Fig. 14), the first
spiro compound isolated from nature with a peculiar spiranic hydantoin
fragment, showed to be a plant growth regulator and a potent herbicide,
without exerting toxicity to humans.91 The synthesis of C-1 spironucleosides
was inspired by hydantocidin 164.
One of the most targeted enzymes for the developing of drugs has been
glycogen phosphorylase (GP), a key enzyme for breaking down glycogen
and thus, a therapeutic target in the treatment of type 2 diabetes. Although
this carbohydrate-proccessing enzyme is not a glycosidase, it deserves some
attention within this review due to its pharmacological importance.
240 | Carbohydr. Chem., 2012, 38, 215–262

O OH OH
NH
O O H O H
HO N HO N
HO NH HO HO
O S
OH OH
O NH NH
HO OH O O
Hydantocidin 164 165 166
Fig. 14
OH
N OAc
OH
OAc
a. NBS, hν
R Cl O CCl4 or CHCl3
AcO S R O
O AcO HO S
AcO HO
AcO SH Et N, CH Cl b. NaOMe, MeOH R
3 2 2
AcO or MeOH, Et3N, H2O
N OH
AcO O N
HO
167 168 169
Scheme 21
OAc OH
N OAc OH
O
R Cl O NaOMe O
AcO HO
AcO CH2 AcO R HO R
AcO Et3N, CH2Cl2
MeOH
AcO OH
O N O N
OAc
170 171 172
Scheme 22
A new type of GP inhibitors that emerged after the isolation of hydan-

tocidin was D-glupopyranosylidene-spiro-hydantoin92 165 (Fig. 14), an
extremely potent competitive GP inhibitor (Table 3), but with limited access
due to a long synthetic pathway.92,93
Somsák et al. described94 the more accessible D-glupopyranosylidene-
spiro-thiohydantoin analogue 166 (Fig. 14), which exhibited good inhibi-
tory profile against GP (Table 3).93 In a recent study,95 it was observed that
166 caused the inhibition of hepatic glycogen breakdown in rats, which
resulted in a rapid decrease of the concentration of glucose in blood.
Crystallographic investigations of GP-inhibitor complexes have recently
been accomplished in order to get insight into mechanistic aspects of inhi-
bition.96 Derivatives like 165 and 166 allowed the development of a series of
new spiranic inhibitors.97 In this context, glycopyranosylidene-spiro-oxa-
thiazoles 169, bearing aromatic residues (remarkable examples are depicted
in Table 3) are noticeable. This family of compounds was obtained98 by
reaction of glucopyranosyl-hydroximothioate derivatives 168 with N-bro-
mosuccinimide (NBS) (Scheme 21). Compounds 169 showed from modest
to excellent inhibition of muscle GPb (Table 3).
Glucopyranosylidene-spiro-isoxazolines 172 were synthesized from the
exo-glucal 170 via selective 1,3-dipolar cycloadition with a series of aromatic
nitrile oxides, generated in situ (Scheme 22).99 Such spiro-isoxazolines
showed moderate to good inhibition of muscle GPb (Table 3);
Carbohydr. Chem., 2012, 38, 215–262 | 241

OBn
O
BnO CH2
BnO
OBn OH
OBn
OBn Et3N NBoc O
173 S BnO
O
HO
O O O BnO HO
BnO NR NR
BnO BnO OH
NHR THF BocN S O HN S O
OBn
BnO
OH O O
O
BnO
BnO
O 175 176 177
OBn
174
Scheme 23
OH OBn OH
O O O
HO HO
HO BnO HO
O BnO CH2 O
OH OH
N S O HN S O
MeO2C OBn
O O
179 173 178
No inhibition Ki 190 μM (α-glucosidase, yeast)
Ki 258 μM (amyloglucosidase,
Aspergillus Niger)
Fig. 15
crystallographic data showed binding of all derivatives to the catalytic site

of the enzyme, via direct and water-mediated hydrogen bonds.
Some glycomimetics that can be obtained from exo-glycals are C-
glycosylated spiro-sulfamides100 and spiro-sulfamidate glycosides.101 The
spiranic sulfamides were obtained starting either from exo-glucal 173, or
from glucoronolactone 174, which were converted into b-aminoalcohols 175
(Scheme 23); treatment of 175 with a Burgess-type reagent furnished the
spiro-sulfamides 176 (Scheme 23).100 The p-anisyl derivative 177a showed
good inhibition against muscle GPb, with a Ki of 15 mM (Table 3).
A similar procedure was used in the synthesis of spiro-sulfamidates
(Fig. 15).101 Compound 178 was tested against a-glucosidase (Ki=190 mM)
and amyloglucosidase (Ki=258 mM); such compound turned out to be only
a moderate inhibitor of both enzymes. Nevertheless, no inhibition was
observed for 179 against any of the tested enzymes, suggesting that the
anomeric NH group is essential for the glycosidase inhibition.
Among all the GP inhibitors described in this chapter, the 2-napthyl
oxathiazol 169b and 2-napthyl isoxazoline 172c turned out to be the two
most potent inhibitors of muscle GPb up to date (Table 3), with inhibition
constants in the nanomolar range. X-Ray crystallographic studies demon-
strated that the 2-naphtyl moiety interacts with several residues of the
enzyme in its less active T-state conformation.99
A series of D-arabino, L-fuco and L-rhamno-configured imino
sugars bearing a spirocyclopropyl fragment have been described by
Behr’s group.102–104 The synthetic pathway involved the formation of
an oxime, followed by subsequent dehydration reaction to afford nitriles
181 or 182 (Scheme 24); the next key steps are a titanium-mediated
242 | Carbohydr. Chem., 2012, 38, 215–262

Table 3 Inhibition of GP (Ki in mM).
Compound Ki muscle GPb Ki liver GPb
OH
O H
HO N
HO 4.2 12.8
O
OH
NH
O
165
OH
O H
HO N
HO 5.1 7.0
S
OH
NH
O
166
OH
O OMe
HO S
HO 8.2 —a
OH
O N
169a
OH
O
HO S
HO 0.16 —a
OH
O N
169b
OH
O OMe
HO
HO 6.6 —a
OH
O N
172a
OH
O Me
HO
HO 7.9 —a
OH
O N
172b
OH
O
HO
HO 0.63 —a
OH
O N
172c
OH
O OCH3
HO
HO
N 15 —a
OH
HN S O
O
177a
a
Value not available or not tested.
Carbohydr. Chem., 2012, 38, 215–262 | 243

OH H H
H N N
N OH OH OH
OMs H
RO 1. EtMgBr, Ti(O-iPr) N
CN HO OH HO OH
–78°C to rt. HO OH
2.BF3 OEt2 RO 186
184 185
RO OR 3.H2O
181 183 H H
N N
OH OH
HO OH HO OH
O
OH
187 188
RO
180 H H
N N
244 | Carbohydr. Chem., 2012, 38, 215–262

HO OH HO OH
CH2OR
CH2OR
RO 1. EtMgBr, Ti(O-iPr) OH OH
CN –78°C to rt. RO
NH2 190 191
2.BF3 OEt2
RO OR N N
RO OR
182 189
HO OH HO OH
192 193
Scheme 24
Table 4 Selected enzymatic inhibition data for imino sugars 184–188 and 190–193.a
Amylogluco-
a-L-Rhamnosidase sidase
a-L-Fucosidase (Penicillium b-Galactosidase (Aspergillus
Compound (bovine kidney) decumbens) (bovine liver) Niger)
OH
H
N
OH
97% (1.6 mM) —b 9% N.I.c
HO OH
184
H
N
OH
50% —b 24% 82% IC50
100 mM
HO OH
186
H
N
83% (18 mM) N.I.c —b N.I.c

HO OH
OH
191
76% 87% (57 mM) —b N.I.c

HO OH
192
a
Percentage of inhibition at 1 mM concentration, Ki values in brackets.
b
Value not available or not tested.
c
No inhibition detected at 1mM.
aminocyclopropanation involving the nitrile group and final intramolecular

cyclization.
Imino sugars 184–188 and 190–193 were tested against a series of
glycosidases, and the most significant results are depicted in Table 4.
Compounds 184 and 191 showed to be potent and selective inhibitors of
a-L-fucosidase (Ki=1.6 mM and 18 mM respectively), while on the contrary,
the L-fuco-configured 192 turned out to be only a weak inhibitor of a-L-
fucosidase, but exhibited a strong affinity for a-L-rhamnosidase (Ki=57
mM). Curiously, arabino-configured 186 showed a moderate inhibition of
amyloglucosidase (IC50=100 mM), but was completely inactive against
other a-glucosidases.102
6 Thio, seleno and carbasugars as glycosidase inhibitors

6.1 Thio and seleno sugars
Thio-sugars emerged as interesting scaffolds for glycosidases inhibitors105
after the discovery of the natural glycosidase inhibitors salacinol106 194 and
Carbohydr. Chem., 2012, 38, 215–262 | 245

OH OH OH OH OH
OH
S OSO3 S OSO3 OH
HO HO
HO OH HO OH
α-glucosidases (Rat intestinal Salacinol 194 Kotalanol 195

disaccharidase)
Maltase Ki 0.31 μM Ki 0.23 μM
Succrase Ki 0.32 μM Ki 0.18 μM
Isomaltase Ki 0.47 μM Ki 1.8 μM
Fig. 16
OH OH
O
O
OBn O
S
BnO
+ S OH O
O O HO
BnO OBn
HO OH
196 197 198
Scheme 25
kotalanol 195107,108 (Fig. 16); both compounds were isolated from Salacia
reticulata and have been used in the Indian Ayurvedic tradicional medicine.
Salacinol and kotalanol share a common zwitterionic sulfonium-sulfate
structure and a polyhydroxylated side chain. These compounds showed
potent inhibition against a series of a-glucosidases (Fig. 16 and Table 5), a
property that can explain the hypoglycemic activity109 of Salacia species.
Analogues to salacinol have been synthesized by coupling a thio(sele-
no)sugar with alkylating species, like epoxide 197 (Scheme 25). In this case,
such coupling afforded the zwitterionic derivative 198,110 a potent inhibitor
of human maltase glucoamylase (MGA) (Table 5).
Following this strategy, several thio(seleno)sugars have been described
by Pinto’s group; these compounds showed potent inhibition of MGA
(Table 5). From these data, it is clear that the presence of a sulfate group at
C-3 of the side chain is essential for a good enzymatic affinity, demonstrated
by the high Ki values of compounds 198, 206 and 207, which lack such
group or bear it at a different position. The sulfate moiety present in some of
the derivatives was introduced by nucleophilic attack of protected five-
membered thio or selenosugars on 1,3,-cyclic sulfates.111 Furthermore, a
change in the configuration of the heteroatom was found to decrease the
potency of the corresponding inhibitor (e.g. 200 vs. 201).
Some bicyclic thiosugars related to swainsonine and 8-epi-lentiginosine
have been synthesized from D-lyxose (Scheme 26).116 The inhibitory
246 | Carbohydr. Chem., 2012, 38, 215–262

Table 5 Enzymatic inhibition of sulfonium-based thiosugars.
MGA (Ki MGA (Ki

Compound in mM) Compound in mM)
OH OH
OH OH
OH
S OSO 105 Se
0.19 OH
0.10112
HO HO
HO OH
HO OH
Salacinol 194
202
OH OH OH OH OH
OH
OH
S 113 Se OH
OH
0.19 0.10112
HO HO
HO OH HO OH
Kotalanol 195 203
OH OH OH OH OH
O OH
S OH O 10110 S OH
0.13114
HO HO
HO OH HO OH
198 204
OH OH OH OH OH
OH
OH
S
S OH 0.65111 OH
0.10114
HO HO
HO OH HO OH
199 205
OH OH OH
OH
OH
Se OH
0.14111 S OH 20115
HO HO
HO OH
HO OH
200 206
OH OH OH
OH
OH
Se OH 10.1111 Se OH 53115
HO HO
HO OH HO OH
201 207
Carbohydr. Chem., 2012, 38, 215–262 | 247

Cl Cl
S S S
HO OH OH
D-Lyxose
BnO OBn OH OH
OH
208 209 210
IC50 14 μM (dGMII) IC50 2000 μM (dGMII)
Scheme 26
OH OH
OH
HO Br OH
HO OH
OH OH
HO
HO OH OH OH
OH OH
211 212 213
IC50 8 μM α-glucosidase IC50 30 μM α-glucosidase IC50 4 μM α-glucosidase
Fig. 17
activities of 8-epi-thiolentiginosine 209 and di-epi-thioswainsonine 210

against Drosophila Golgi a-mannosidase II (dGMII) were tested. Com-
pound 209 showed a much stronger inhibition (IC50=14 mM) than deri-
vative 210 (IC50=2000 mM).
6.2 Carbasugars
In the literature, the synthesis of carbasugars is considered as a useful source
of glycosidase inhibitors,117 like polihydroxylated derivatives 211–213 (Fig.
17).118–120
Conduramine F-1 epoxides 216–218 (Scheme 27)121 were synthesized
starting from ()-7-oxanorbornenone 214, using cyclohexene derivative
215 as the key intermediate, which was subjected to an epoxidation reaction.
All compounds behaved as poor inhibitors against b-D-galactosidase
(bovine liver) and amyloglucosidase (Aspergillus niger); however com-
pounds 216 and 217, bearing trans epoxy and amino moieties, exhibited
potent non-competitive inhibition of b-D-xylosidase (Aspergillus niger;
Ki=48 mM and 2.2 mM, respectively). On the other hand, compound 218,
with a cis arrangement of the epoxy and amino functionalities, turned out to
be a non competitive strong inhibitor of a-glucosidase (brewer’s yeast) with
a Ki value of 2.8 mM. It is also remarkable that non-substituted derivatives
on the nitrogen atom were found to be quite less active.
The structure of the well known carbasugar valienamine has also been an
inspiration for the development of new glycosidase inhibitors.117 6-epi-
Valienamine derivative 222, bearing an acetamido group at C-2 of the
pseudo-sugar ring, was synthesized from protected methyl glucopyranoside
219 in a 14-step synthetic pathway (Scheme 28).122 Compound 222 showed
248 | Carbohydr. Chem., 2012, 38, 215–262

O O
OH OH
OH HN OH HN
OR
O OH OH OH CF3
RO N3
O 216 217
Ki 48 μM β-xylosidase (A. niger) Ki 2.2 μM β-xylosidase (A. niger)
RO O
(±) 214 215
OH
OH HN
OH
218
Ki 2.8 μM α-glucosidase (brewer's yeast)
Scheme 27
OH
Ph O HO
O
O O
BnO BnO BnO HO
BnO OBn
BnO HO NH3 Cl
AcHN
OMe N3 N3 AcHN
OMe
219 220 221 222

Ki 50 μM family 3 β-N-acetyl-glusamidase (vibrio cholerae)
Ki 34 μM family 20 β-hexosaminidase (human placenta)
Ki 6.2 μM family 84 O-GlcNAcase (human)
Scheme 28
Table 6 Inhibitory activity of isoureas and guanidines derived from carbasugars (IC50 in mM).
OH
OH OH
HO
HO NH HO NH HO
HO HO NH
O
NHR HO NHR
HO O HO
NH NHR
225
Imiglu-
cerase
a () b () c () a () b () c () a (–) b (–)

pH 7 1.6 0.08 0.04 0.04 0.01 0.008 0.004 0.01
pH 5.2 6.4 0.60 0.22 0.10 0.02 0.009 0.01 0.04
strong inhibition of human b-hexosamidase (Ki=34 mM) and human O-

GlcNAcase (Ki=6.2 mM).
A series of N-alkylated bicyclic isoureas and guanidines 223–225 were
recently prepared from 1,4-aminoalcohols or 1,4-diamines; the key steps are
a HgO-promoted cyclodesulfurization reaction of a thiourea-type inter-
mediate to give a transient carbodiimide, followed by cyclization involving
the hydroxyl or amino groups located at the C-4 position.123 These com-
pounds showed a remarkable inhibition of recombinant b-glucocer-
ebrosidase (imiglucerase) in the nM range, and in most cases in a pH
dependent-manner (Table 6).
Carbohydr. Chem., 2012, 38, 215–262 | 249

7 Miscellaneous
Although imino sugars, azasugars and related compounds are the most
frequent inhibitors of carbohydrate-processing enzymes, there are some
other different architectures that also show valuable inhibitory properties.
In this context, Withers and co-workers accomplished the synthesis of
thioglycosides as non cleavable substrate analogues designed for the inhi-
bition of glycosidases and carbohydrate-binding proteins.124 In particular, a
small library of galactosyl thioglycosides 228–230 was accessed by coupling
2,4-dinitrophenyl b-D-galactopyranoside (226) with thiol-containing car-
bohydrates of D-gluco and D-galacto configurations (227); such reaction was
promoted by a thioglycoligase derived from the Xanthomonas manihotis
b-galactosidase (BgaX) (Scheme 29). Title compounds were tested against a
series of b-galactosidases. Gal-S-Glc derivative 228 showed to be a potent
OH
OH
NO2
O O
O thioglycoligase
HO
+ 228-230
HS
OH
OR
NO2
226 227
β-Galactosidase (Ki)
E. Coli X. maihotis human lysosomal
OH
OH
OH
O 9 μM 700 μM 8 μM
O HO O
HO S
OH
OH
NO2
228 (85%)
OH
OH HO
OH
O
O O
HO S 51 μM 1200 μM 350 μM
OH
OH
NO2
229 (83%)
OH
OH
OH
OH
O
150 μM 1300 μM 1700 μM
HO O
S HO O
OH
NO2
230 (80%)
Scheme 29
250 | Carbohydr. Chem., 2012, 38, 215–262

enzyme-inactivator
OAc OH covalent complex
OH
O F
O F
HO α-glucosidase O O
AcO HO HO
AcO HO
O
Cl Cl
F Cl
231 232 (ki/Ki 0.0053 min–1 mM–1) 233
Scheme 30
inhibitor of human and E. coli galactosidases. An inversion of the stereo-

chemistry of C-4 of the terminal sugar unit (229) was found to decrease
about 40 times the inhibition; furthermore, even lower affinity was observed
by changing the linkage from b-S-1,3 to b-S-1,4 (230) (Scheme 29).124
Withers also designed125 a series of 2-deoxy-2,2-dihalo glucosyl chlorides
as a useful class of mechanism-based inhibitors of a-glucosidase; the pre-
sence of a leaving group at the anomeric position allowed trapping glycosyl-
enzyme intermediates, what can be of interest as potential therapeutics.
In this context, 2-chloro-2-deoxy-2-fluoro-a-D-glucosyl chloride derivative
(232) was synthesized by a chlorination reaction of 2-fluoroglucal 231
(Scheme 30). Compound 232 was found to be a time-dependent inactivator
of a-glucosidase (yeast), as covalent glycosyl-enzyme intermediate 233 could
be obtained upon interaction with the enzyme.
Cu(I)-Catalyzed alkyne-azide cycloaddition has become a useful reaction
for linking sugar and triazolyl scaffolds in a regioselective fashion. In this
context, a series of 4-substituted 1,2,3-triazolyl-sugar derivatives have
been synthesized as potential glycosidases inhibitors.126,127 Among the
compounds that bear a triazolyl moiety at the anomeric position, it has
been demonstrated that a-configured triazolyl derivatives inhibit yeast
a-glucosidase preferentially over their b-anomers; however the inhibition
observed was found to be modest.127 By changing the triazolyl scaffold to
the positions 3 or 5 of the sugar ring (Fig. 18),128,129 the corresponding
derivatives were found to be moderate to good inhibitors of glycosidases,
despite the carbohydrate residues were kept fully protected. Especially
noticeable are derivatives 237 and 238, which were reported to exhibit a
potent inhibition of yeast maltase (IC50=3.8 and 5.2 mM, respectively)
(Fig. 18).129
8 Pharmacological activities of imino sugars

Imino sugars are characterized by a high pharmaceutical potential; up to
date, hundreds of different imino sugars and related compounds have been
reported to exhibit strong interactions with an ample number of essential
glycosidases and other carbohydrate-processing enzymes. Therefore, such
carbohydrate mimics can address a vast number of biological targets, being
candidates for the treatment of diseases ranging from cancer to lysosomal
storage disorders or microbial infections, among others (Fig. 19).130,131
Inspired by naturally-occurring imino sugars,132 a great number of syn-
thetic imino sugars emerged in the so-called first generation. Although
Carbohydr. Chem., 2012, 38, 215–262 | 251

O
N O
O N
Ph
N O N
N
Ph O
N OH
O
OH
Ph
234 235
IC50 119.5 μM α-glucosidase (Saccaromyces cerevisiae) IC50 247.4 μM α-glucosidase (Saccaromyces cerevisiae)
O
O
O OMe
O N
N
O O N OMe
N N
N
O O
N N O
O O
N Ph
Ph
236 237 238
99.4% of inhibition at 500 mM, 98.1% of inhibition at 500 mM, 98.5% of inhibition at 500 mM,
α-glucosidase (yeast) α-glucosidase (yeast) α-glucosidase (yeast)
IC50 3.8 μM maltase (yeast) IC50 5.2 μM maltase (yeast)
Fig. 18
many of them lacked enzymatic selectivity, two drugs were approved and
marketed for clinical use, and some others are in advanced clinical-stages:
Miglitols (N-2-hydroxyethyl-deoxynojirimycin) and Zavescas (N-butyl-
deoxynojirimycin) are currently used for the treatment of type 2 diabetes
and Gaucher/Niemann-Pick type C diseases (substrate reduction therapy),
respectively.130
8.1 Imino sugars for the treatment of lysosomal storage disorders

Lysosomal storage disorders are diseases caused by the impairment of
lysosomal enzymes, leading to the accumulation of partially degraded
metabolites in the intra-lysosomal region of several organs. Gaucher
(accumulation of glucosylceramide), Fabry (accumulation of globo-
triaosylceramide), Tay-Sachs, and Sandhoff (accumulation of N-acetyl-
hexosamine-terminating glycolipids or N-glycoproteins, respectively)
diseases are remarkable examples.1,133,134 In some of the lysosomal storage
disorders, handicapped proteins are misfolded, but still keep some catalytic
activity; nevertheless, they are targeted and degraded by the quality control
machinery on the endoplasmatic reticulum. One promising therapy for the
treatment of such diseases is called the chaperone therapy, where a glyco-
sidase inhibitor at sub-inhibitory concentrations binds the active site of the
enzyme and induces the proper folding.1
Gaucher disease is the lysosomal storage disorder most frequently stu-
died; an ample number of imino sugar derivatives have been designed and
proposed for chaperoning-based studies, ranging from unsubstituted/
N-alkylated deoxynojirimycins,135 bicyclic guanidines,136 noeurostegines,137
252 | Carbohydr. Chem., 2012, 38, 215–262

OH O
N
HO (CH2)6NH-Dansyl
HO N
OH
OH (H2C)6
NH-Dansyl
HO
N O
239
Ki 1.7 μM (lysosomal β-glucosidase)
HO
Chemical chaperones
for lysosomal storage OH
disorders 33
Antidiabetic and anti-obesity agents
HO
O
OH O U
i N P O
N O Pr CH3
HO
HO 6 HO O
15
CH3 HO OH
240 241
EC50 1.6 μM (bovine viral diarrhea virus) MIC 1.56 μg/mL (S. agalactinae)
Lipophosphonoxins
Antimicrobial agents
(CH2)9CH3
OH OH
H
N
O
HO
OH HO
N
N S
HO OH
D-Lentiginosine 242 NH
244
Anticancer agents
Immunosuppresants
Thiamet-G 243
Ki 21 nM (O-GlcNAcase)
Potential treatment of
neurodegerative diseases
Fig. 19
N-substituted d-lactams138 or tetrahydroxyazepanes,139 among others.

Stütz’s group has developed140 a series of fluorescently-tagged deoxynojir-
imycins bearing dansyl groups for diagnosis purpose in lysosomal storage
disorders. For instance, compound 239 (Fig. 19), a potent inhibitor of
lysosomal b-glucosidase (Ki=1.7 mM), was effective at enhancing mutant
lysosomal glucocerebrosidase activity.
1-Deoxygalactonojirimycin (migalastat) was found141 to increase a-
galactosidase A activity in mice upon oral administration, and also to sig-
nificantly decrease the concentration of globotriaosylceramide in skin,
heart, kidney and even in brain, as this compound was capable of crossing
the blood-brain barrier. These results might suggest a potential therapeutic
use in Fabry disease.
N-Butyldeoxynojirimycin, already used in clinical treatment of Gaucher
disease, was chosen142 as a model compound for studying Sandhoff disease in
mice; surprisingly, it was found that this compound was efficient for lowering
the concentration of forebrain ganglioside content, and the combination with a
restricted ketogenic diet enhanced its transportation to brain.
Carbohydr. Chem., 2012, 38, 215–262 | 253

8.2 Antidiabetic agents
Type 1 and 2 diabetes mellitus are a worldwide health concern provoking
chronic hyperglycemia, which in turn leads to long-term damage in several
organs and cardiovascular diseases. Two ways of combating type 2 diabetes
using enzymatic inhibitors can be suggeted:1 to regulate digestive a-gluco-
sidases, or to decrease the hepatic biosynthesis of glucose mediated by
glycogen phosphorylase143 (e.g. by using spiranic compounds covered in
section 5.3).
In this context, research on imino sugar-based glycosidase inhibitors led
to miglitol, an intestinal a-1,4-glucosidase inhibitor, marketed by the Bayer
Company in 1996 for use in the United States, South America and some
European countries.144
There are evidences that correlate obesity and the excess of lipidic matter
with diminished response to insulin.145 Recently, attention has been focused
on lipophilic N-[5-(adamantan-1-ylmethoxy)pentyl]-1-deoxynojirimycin 33
(AMP-DNM, Fig. 4) which inhibits both, the synthesis of glycosylceramide
(the precursor of glycosphingolipids like glangliosides), and carbohydrate
assimilation, which in turn enhances hepatic insulin sensitivity.146,147
Langeveld et al. suggested148 that this compound could be a good candidate
for the treatment of obesity and metabolic diseases, as AMP-DNM effi-
ciently reduced the weight of genetically obese mice by increasing fat oxi-
dation and by reducing the animal food intake. Furthermore, it is also a
good candidate for improving glucose homeostasis and hepatic steatosis,
the first stage of non-alcoholic liver disease.149
Moreover, AMP-DNM also provoked a deep decrease in the amount of
lipids present in plasma by stimulating reverse cholesterol transportation;
this proved to protect mice against atherosclerosis development.150
8.3 Imino sugars as antimicrobial agents

It is well-documented that a way of combating viral infections on humans
(e.g. HIV, hepatitis B and C, or influenza) is to affect the viral N-glycan
processing pathway; a-glucosidases are one of the enzymes involved in this
process.144 This observation led to the identification of some N-alkylated
imino sugars as good antiviral agents like N-nonyl-deoxynojirimycin, but
with usually poor cytotoxic profiles. In this context, Chang et al. have
improved151 the therapeutic profile of N-alkylated deoxynojirimycins
against viruses of the Flaviviridae family like BVDV (bovine viral diarrhea
virus) and DENV (dengue virus). They found that an oxygenated alkyl side
chain bearing a terminal ring, either cycloalkyl or aromatic, was crucial for
maintaining a strong antiviral activity, while low citotoxicity. In particular,
compound 240 (Fig. 19) was found to inhibit in vitro DENV infection with
90% effective concentration at submicromolar concentrations.
Regarding antibacterial derivatives, Krásný and co-workers have recently
reported152 a novel family of imino sugars named lipophosphonoxins 241,
bearing four different molecular modules: an imino sugar, a nucleoside
scaffold, a lipophilic alkyl chain module and a phosphonate linker (Fig. 19).
The optimized structure-activity relationships of these four structural motifs
afforded compounds exhibiting strong antibacterial activity against a panel
of Gram-positive species, including multiresistant strains. Title compounds
254 | Carbohydr. Chem., 2012, 38, 215–262

exhibited good MIC (Mininum Inhibitory Concentration) values, while
higher cytotoxic concentrations against human cell lines.
8.4 Imino sugars as antitumor agents

It has been reported that one of the processes that accompanies the growth
of tumour cells is aberrant glycosylation, where glycosidases are obviously
involved.153 In the imino sugars field, some glycosidase inhibitors have been
found to possess antitumoral activity; the potential use of imino sugars as
anticancer agents was recently reviewed by Wrodnigg et al.154 The best-
known imino sugars exhibiting antitumoral activity are castanospermine
and swainsonine; the latter was included in phase I clinical studies144 for
treating advanced tumours, as it prevented metastatic growth. Nevertheless,
the significant hepatoxicity of this compound precluded any further study.
Therefore, deeper search for new imino sugar-based templates is required
for developing new antitumoral agents with better pharmacological profiles.
In this context, Xiao’s group reported155 the successful use of a microtiter-
plate based method applied to live cells for screening antitumor activity
using a library of imino sugar derivatives.
Inspired by the indolizidine-type alkaloids castanospermine and swain-
sonine, Macchi and co-workers focused attention on lentiginosine, a potent
amyloglucosidase inhibitor. This group found156 that the non-natural
enantiomer (D-lentiginosine 242, Fig. 19) induced caspase-dependent
apoptosis on tumor cells at relatively low concentrations, in contrast with its
natural enantiomer.
Promising anticancer activity has also been found in pyrrolidine-157 and
piperidine-157a,158 based motifs. Well-known natural 1-deoxynojirimycin
was found to inhibit the metastasis of B16F10 melanoma cells; studies on its
antimetastatic mechanism suggested that this compound could be a good
candidate as adjuvant of other antimestatic agents, like cisplatin.158 Padró
et al. accomplished a structure-activity study for N-alkylated D-fagomine
and a trihydroxylated pyrrolidine, which revealed a direct correlation
between the length of the alkyl residue and the citotoxicity in human cancer
cell lines.157a
8.5 Treatmet of neurodegenerative diseases

Vocadlo’s group reported159 the synthesis of thiamet-G 243 (Fig. 19), a cis-
fused bicyclic thiazolidine, the hitherto most potent inhibitor of eukaryotic
O-GlcNAcases (Ki=21 nM). This compound behaved as a highly selective
inhibitor, as had no effect on b-hexosaminidases. Thiamet-G was proved to
increase cellular levels of O-GlcNAc in neuron-like PC-12 cells, which in
turn blocks phosphorylation of microtubule-associated protein tau at
Thr231, Ser396 and Ser422; pathological hyperphosphorylation of two of
these sites is characteristic in neurodegenerative Alzheimer’s disease and
associated taupathies.
8.6 Other biological activities of imino sugars

Ye and co-workers described160 new N-alkylated 1,5-dideoxy-1,5-imino-
pentitols and N-alkylated 1,4-dideoxy-1,4-iminotetritols (e.g. 244, Fig. 19)
which inhibited the mouse splenocyte proliferation by suppressing IFN-g
Carbohydr. Chem., 2012, 38, 215–262 | 255

and IL4-cytokines. As a result, such compounds prolonged the allograft
survival in a mouse skin transplantation experiment, that is, they exhibited
immunosuppressive activities, the first synthetic imino sugars showing such
property in animal models.
9 Concluding remarks
Imino sugars, despite being studied for more than 40 years, are still a main
goal in Medicinal Chemistry. Nature is a continuous source of new imino
sugar templates, which in turn are the inspiration of synthetic chemists to
design novel families with improved activity, bioavailability and selectivity.
The broad pharmacological spectrum of this sugar mimics justify the cur-
rent intense research in this area.
Acknowledgements
We thank the Dirección General de Investigación of Spain (CTQ2008
02813) and Junta de Andalucı́a (FQM 134) for financial support. S. M.
thanks MICINN of Spain for the award of a fellowship.
References
1 M. Bols, Ó. López, and F. Ortega-Caballero, in Comprehensive Glycoscience-
From Chemistry to Systems Biology, Ed. J.P. Kamerling, Elsevier, 2007, 815.
2 T. M. Gloster and G. J. Davies, Org. Biomol. Chem., 2010, 8, 305.
3 J. Wicki, S. J. Williams and S. G. Withers, J. Am. Chem. Soc., 2007, 129, 4530.
4 M. van Scherpenzeel, R .J. B. H. N. van der Berg, W. E. Donker-Koopman,
R. M. J. Liskamp, J. M. F. G. Aerts, H. S. Overkleeft and R. J. Pieters, Bioorg.
Med. Chem., 2010, 18, 267.
5 T. M. Gloster, P. Meloncelli, R. V. Stick, D. Zechel, A. Vasella and G. J.
Davies, J. Am. Chem. Soc., 2007, 129, 2345.
6 W. A. Offen, D. L. Zechel, S. G. Withers, H. J. Gilbert and G. J. Davies,
Chem. Commun., 2009, 2484.
7 A. C. Sivertsen, M. Gasior, M. Bjerring, S. U. Hansen, Ó. López López, N. C.
Nielsen and M. Bols, Eur. J. Org. Chem, 2007, 1735.
8 M. E. C. Caines, S. M. Hancock, C. A. Tarling, T. M. Wrodnigg, R. V. Stick,
A. E. Stütz, A. Vasella, S. G. Withers and N. C. J. Strynadka, Angew. Chem.
Int. Ed., 2007, 46, 4474.
9 Y.-X. Li, M.-H. Huang, Y. Yamashita, A. Kato, Y.-M. Jia, W.-B. Wang,
G. W. J. Fleet, R. J. Nash and C.-Y. Yu, Org. Biomol. Chem., 2011, 9, 3405.
10 D. Best, S. F. Jenkinson, A. W. Saville, D. S. Alonzi, M. R. Wormald, T. D.
Butters, C. Norez, F. Becq, Y. Blériot, I. Adachi, A. Kato and G. W. J. Fleet,
Tetrahedron Lett., 2010, 51, 4170.
11 I. Izquierdo, M. T. Plaza, J. A. Tamayo, F. Sánchez-Cantalejo and D. L. Re,
Synthesis, 2008, 1373.
12 E. Moreno-Clavijo, A. T. Carmona, Y. Vera-Ayoso, A. J. Moreno-Vargas,
C. Bello, P. Vogel and I. Robina, Org. Biomol. Chem., 2009, 7, 1192.
13 J. S. S. Rountree, T. D. Butters, M. R. Wormald, S. D. Boomkamp, R. A.
Dwek, N. Asano, K. Ikeda, E. L. Evinson, R. J. Nash and G. W. J. Fleet,
Chem. Med. Chem., 2009, 4, 378.
14 J.-B. Behr, A. Kalla, D. Harakat and R. Plantier-Royon, J. Org. Chem., 2008,
73, 3612.
256 | Carbohydr. Chem., 2012, 38, 215–262

15 A. Kato, J. Hollinshead, Y. Yamashita, S. Nakagawa, Y. Koike, I. Adachi,
C.-Y. Yu, G.W.J. Fleet and R. J. Nash, Phytochem. Lett, 2010, 3, 230.
16 E. D. Goddard-Borger, R. Carapito, J.-M. Jeltsch, V. Phalip, R. V. Stick and
A. Varrot, Chem. Commun., 2011, 47, 9684.
17 Y. Du, Q. Pan and F. Kong, Carbohydr. Res., 2000, 329, 17.
18 V. H. Lillelund, H. H. Jensen, X. Liang and M. Bols, Chem. Rev, 2002, 102,
515.
19 Z. Yu, A. R. Sawkar, L. J. Whalen, C.-H. Wong and J. W. Kelly, J. Med.
Chem., 2007, 50, 94.
20 T. Wennekes, R. J. B. H. N. van der Berg, W. Donker, G. A. van der Marel,
A. Strijland, J. M. F. G. Aerts and H. S. Overkleeft, J. Org. Chem., 2007, 72,
1088.
21 G. Schitter, E. Scheucher, A. J. Steiner, A. E. Stütz, M. Thonhoefer, C. A.
Tarling, S. G. Withers, J. Wicki, K. Fantur, E. Paschke, D. J. Mahuran, B. A.
Rigat, M. Tropak and T. M. Wrodnigg, Beilstein J. Org. Chem, 2010, 6, 21.
22 A. J. Steiner, A. E. Stütz, C. A. Tarling, S. G. Withers and T. M. Wrodnigg,
Carbohydr. Res., 2007, 342, 1850.
23 A. J. Steiner, G. Schitter, A. E. Stütz, T. M. Wrodnigg, C. A. Tarling, S. G.
Withers, K. Fantur, D. Mahuran, E. Paschke and M. Tropak, Bioorg. Med.
Chem., 2008, 16, 10216.
24 A. J. Rawlings, H. Lomas, A. W. Pilling, M. J.-R. Lee, D. S. Alonzi, J. S.
Rountree, J. Shane, G. W. J. Fleet, R. A. Dwek, J. H. Jones and T. D. Butters,
ChemBioChem, 2009, 10, 1101.
25 N. Ardes-Guisot, D. S. Alonzi, G. Reinkensmeier, T. D. Butters, C. Norez, F.
Becq., Y. Shimada, S. Nakagawa, A. Kato, Y. Blériot, M. Sollogoub and B.
Vauzeilles, Org. Biomol. Chem., 2011, 9, 5373.
26 H. S. Overkleeft, G. H. Renkema, J. Neele, P. Vianello, I. O. Hung, A.
Strijland, A. M. von der Burg, G.-J. Koomen, U. K. Pandit and J. M. F. G.
Aerts, J. Biol. Chem., 1998, 273, 26522.
27 J. D. Diot, I. G. Moreno, G. Twigg, C. O. Mellet, K. Haupt, T. D. Butters,
J. Kovensky and S. G. Gouin, J. Org. Chem., 2011, 76, 7757.
28 R. W. Wang, X. L. Qiu, M. Bols, F. Ortega-Caballero and F. L. Quing,
J. Med. Chem., 2006, 49, 2989.
29 R.-j. Li, M. Bols, C. Rousseau, X.-g. Zhang, R.-w. Wang and F.-L. Qing,
Tetrahedron, 2009, 65, 3717.
30 Y. Yang, F. Zheng, M. Bols, L. Marinescu and F.-L. Qing, J. Fluor. Chem,
2011, 132, 838.
31 S. R. S. Ting, G. Chen and Martina H. Stenzel, Polym. Chem, 2010, 1, 1392.
32 T. Wennekes, R. J. B. H. N. van den Berg, K. M. Bonger, W. E. Donker-
Koopman, A. Ghisaidoobe, G. A. van der Marel, A. Strijland, J. M. F. G.
Aerts and H. S. Overkleeft, Tetrahedron: Asymmetry, 2009, 20, 836.
33 J. Diot, M. I. Garcı́a-Moreno, S. G. Gouin, C. Ortiz Mellet, K. Haupt and J.
Kovensky, Org. Biomol. Chem., 2009, 7, 357.
34 P. Compain, C. Decroocq, J. Iehl, M. Holler, D. Hazelard, T. Mena Barragán,
C. Ortiz Mellet and J.-F. Nierengarten, Angew. Chem. Int. Ed., 2010, 49, 5753.
35 C. Decroocq, D. Rodrı́guez-Lucena, V. Russo, T. Mena Barragán, C. Ortiz
Mellet and P. Compain, Chem. Eur. J, 2011, 17, 13825.
36 E. Lindbäck, Ó. López, J. G. Fernández-Bolaños, S.P.A. Sauer and M. Bols,
Org. Lett., 2011, 13, 2908.
37 E. Lindbäck, Y. Zhou, Ó. López, J. G. Fernández-Bolaños, C. M. Pedersen
and M. Bols, Tetrahedron Lett., 2012, 53, 2045.
38 Ó. López López and M. Bols, ChemBioChem, 2007, 8, 657.
39 A. Lohse, K. B. Jensen and M. Bols, Tetrahedron Lett., 1999, 40, 3033.
Carbohydr. Chem., 2012, 38, 215–262 | 257

40 X. Zhu, K. A. Sheth, S. Li, H. H. Chang and J. Q. Fan, Angew. Chem. Int. Ed.
Engl., 2005, 44, 7450.
41 B. V. Ernholt, I. B. Thomsen, A. Lohse, I. W. Plesner, K. B. Jensen, R. G.
Hazell, X. Liang, A. Jakobsen and M. Bols, Chem. Eur. J, 2000, 6, 278.
42 H. H. Jensen, L. Lyngbye, A. Jensen and M. Bols, Chem. Eur. J, 2002, 8, 1218.
43 M. J. Alves, F. T. Costa, V. C. M. Duarte, A. G. Fortes, J. A. Martins and
N. M. Micaelo, J. Org. Chem., 2011, 76, 9584.
44 X. Liang and M. Bols, J. Org. Chem., 1999, 64, 8485.
45 E. Moreno-Clavijo, A. T. Carmona, A. J. Moreno-Vargas, M. A. Rodrı́guez-
Carvajal and I. Robina, Bioorg. Med. Chem., 2010, 18, 4648.
46 H. H. Jensen, A. Jensen, R. G. Hazzel and M. Bols, J. Chem. Soc. Perkin
Trans, 2002, 1, 1190.
47 F. Morı́s-Varas, X.-H. Qian and C.-H. Wong, J. Am. Chem. Soc., 1996, 118,
7647.
48 H. Li, Y. Zhang, P. Vogel, P. Sinaÿ and Y. Blériot, Chem. Commun., 2007,
183.
49 J. Deschamp, M. Mondon, S. Nakagawa, A. Kato, D. S. Alonzi, T. D.
Butters, Y. Zhang, M. Sollogoub and Y. Blériot, Bioorg. Med. Chem., 2012,
20, 641.
50 H. Li, T. Liu, Y. Zhang, S. Favre, C. Bello, P. Vogel, T. D. Butters, N. G.
Oikonomakos, J. Marrot and Y. Blériot, ChemBioChem, 2008, 9, 253.
51 H. Li, F. Marcelo, C. Bello, P. Vogel, T. D. Butters, A. P. Rauter, Y. Zhang,
M. Sollogoub and Y. Blériot, Bioorg. Med. Chem., 2009, 17, 5598.
52 F. Marcelo, Y. He, S. A. Yuzwa, L. Nieto, J. Jiménez-Barbero, M. Sollogoub,
D. J. Vocadlo, G. D. Davies and Y. Blériot, J. Am. Chem. Soc., 2009, 131,
5390.
53 D.M.F. van Aalten, D. Komander, B. Synstad, S. Gaseidnes, M. G. Peter,
and V.G.H. Eijsink, Proc. Natl. Acad. Sci. U.S.A., 2001, 98, 8979.
54 I. Tews, A. Perrakis, A. Oppenheim, Z. Dauter, K. S. Wilson and C. E.
Vorgias, Nat. Struct. Biol., 1996, 3, 638.
55 N. B. Klamkar, V. M. Kasture, S. T. Chavan, S. G. Sabharwal and D. D.
Dhavale, Tetrahedron, 2010, 66, 8522.
56 T.-L. Shih, R.-Y. Yang, S.-T. Li, C.-F. Chiang and C.-H. Lin, J. Org. Chem.,
2007, 72, 4258.
57 T.-L. Shih, M.-T. Liang, K.-D. Wu and C.-H. Lin, Carbohydr. Res., 2011, 346, 183.
58 B. Krämer, T. Franz, S. Picasso, P. Pruschek and V. Jäger, Synlett, 1997, 295.
59 G. M. J. Lenagh-Snow, N. Araujo, S. F. Kenkinson, C. Rutherford, S.
Nakagawa, A. Kato, C.-Y. Yu, A. C. Weymouth-Wilson and G. W. J. Fleet,
Org. Lett., 2011, 13, 5834.
60 J. P. Michael, Nat. Prod. Rep., 2008, 25, 139.
61 B. La Ferla, L. Cipolla, and F. Nicotra, in Iminosugars-from Synthesis to
Thereapeutic Applications, Ed. P. Compain and O.R. Martin, John Wiley &
Sons, 2007, 25.
62 L. Cipolla, M. R. Fernandes, M. Gregori, C. Airoldi and F. Nicotra, Carbo-
hydr. Res., 2007, 342, 1813.
63 A. Michalik, J. Hollinshead, L. Jones, G. W. J. Fleet, C.-Y. Yu, X.-G. Hu,
R. van Well, G. Horne, F. X. Wilson, A. Kato, S. F. Jenkinson and R. J. Nash,
Phytochem. Lett, 2010, 3, 136.
64 X.-G. Hu, B. Bartholomew, R. J. Nash, F. X. Wilson, G.W.J. Fleet, S.
Nakagawa, A. Kato, Y.-M. Jia, R. van Well and C.-Y Yu, Org. Lett., 2010, 12,
2562.
65 M. Aguilar-Moncayo, C. O. Mellet, J. M. G. Fernández and M. I. Garcı́a-
Moreno, J. Org. Chem., 2009, 74, 3595.
258 | Carbohydr. Chem., 2012, 38, 215–262

66 M. Aguilar-Moncayo, P. Dı́az-Pérez, J. M. Garcı́a Fernández, C. Ortiz Mellet
and I. Garcı́a Moreno, Tetrahedron, 2012, 68, 681.
67 M. Aguilar-Moncayo, T. M. Gloster, J. P. Turkenburg, I. Garcı́a Moreno, C.
Ortiz Mellet, G. J. Davies and J. M. Garcı́a Fernández, Org. Biomol. Chem.,
2009, 7, 2738.
68 B. Brumshtein, M. Aguilar-Moncayo, I. Garcı́a Moreno, C. Ortiz Mellet,
J. M. Garcı́a Fernández, I Silman, Y. Shaaltiel, D. Aviezer, J. L. Sussman and
A. H. Futerman, ChemBioChem, 2009, 10, 1480.
69 E. M. Sánchez-Fernández, R. Rı́squez-Cuadro, M. Chasseraud, A. Ahidouch,
C. Ortiz Mellet, H. Ouadid-Ahidouch and J. M. Garcı́a. Fernández, Chem.
Commun., 2010, 46, 5328.
70 G. Pandey, D. Grahacharya, K. S. Shashidhara, M. I. Khan and V. G.
Puranik, Org. Biomol. Chem., 2009, 7, 3300.
71 F. Cardona, C. Parmeggiani, E. Faggi, C. Bonaccini, P. Gratteri, L. Sim,
T. M. Gloster, S. Roberts, G. J. Davies, D. R. Rose and A. Goti, Chem. Eur. J,
2009, 15, 1627.
72 C. Bonaccini, M. Chioccioli, C. Parmeggiani, F. Cardona, D. Lo, Re, G.
Soldaini, P. Vogel, C. Bello, A. Goti and P. Gratteri, Eur. J. Org. Chem, 2010, 5574.
73 Ó. López, J. G. Fernández-Bolaños and M. V. Gil, Green Chem., 2005, 7, 431.
74 N. Asano, H. Kuroi, K. Ikeda, H. Kizu, Y. Kameda, A. Kato, I. Adachi, A. A.
Watson, R. J. Nash and G.W.J. Fleet, Tetrahedron, Asymmentry, 2000, 11, 1.
75 a) L. Chabaud, Y. Landais, and P. Renaud, Org. Lett., 2005, 7, 2587;
b) S. Chandrasekhar, B.B. Parida, and C. Rambabu, J. Org. Chem., 2008,
73, 7826; c) P.V. Reddy, A. Veyron, P. Koos, A. Bayle, A.E. Greene, and P.
Delair, Org. Biomol. Chem., 2008, 6, 1170; d) T. J. Donohoe, R. E. Thomas,
M. D. Cheeseman, C. L. Rigby, G. Bhalay, and I. D. Linney, Org. Lett., 2008,
10, 3615.
76 G. D’Adamio, A. Goti, C. Parmeggiani and E. Moreno-Clavijo, I. Robina,
and F. Cardona, Eur. J. Org. Chem, 2011, 7155.
77 T. S. Rasmussen and H. H. Jensen, Org. Biomol. Chem., 2010, 8, 433.
78 T. M. Gloster, R. Madsen and G. J. Davies, ChemBioChem, 2006, 7, 738.
79 D. L. Zechel, A. B. Boraston, T. Gloster, C. M. Boraston, J. M. Mcdonald,
M. G. Tilbrook, R. V. Stick and G. J. Davies, J. Am. Chem. Soc., 2003, 125,
14313.
80 H. Liu, X. Liang, H. Søhoel, A. Bülow and M. Bols, J. Am. Chem. Soc., 2001,
123, 5116.
81 M. Aguilar, T. M. Gloster, I. Garcı́a Moreno, C. Ortiz Mellet, G. J. Davies, A.
Llebaria, J. Casas, M. Egido-Gabás and J. M. Garcı́a Fernández, Chem-
BioBhem, 2008, 9, 2612.
82 V. Chagnault, P. Compain, K. Lewinski, K. Ikeda, N. Asano and O. R.
Martin, J. Org. Chem., 2009, 74, 3179.
83 S. Knapp, D. Vocadlo, Z. N. Gao, B. Kirk, J. P. Lou and S. G. Withers, J. Am.
Chem. Soc., 1996, 118, 6804.
84 G. E. Whitworth, M. S. Macauley, K. A. Stubbs, R. J. Dennis, E. J. Taylor, G.
J. Davies, I. R. Greig and D. J. Vocadlo, J. Am. Chem. Soc., 2007, 129, 635.
85 S. Knapp, D. Fash, M. Abdo, T. J. Emge and P. R. Rablen, Bioorg. Med.
Chem., 2009, 17, 1831.
86 S. Knapp, M. Abdo, K. Ajayi, R. A. Huhn, T. J. Emge, E. J. Kim and J. A.
Hanover, Org. Lett., 2007, 9, 2321.
87 J. M. Macdonald, C. A. Tarling, E. J. Taylor, R. J. Dennis, D. S. Meyers,
S. Knapp, G. J. Davies and S. G. Withers, Angew. Chem. Int. Ed., 2010, 49, 2599.
88 B. Amorelli, C. Yang, B. Rempel, S. G. Withers and S. Knapp, Bioorg. Med.
Chem., 2008, 18, 2944.
Carbohydr. Chem., 2012, 38, 215–262 | 259

89 E. J. Kim, D. C. Love, E. Darout, M. Abdo, B. Rempel, S. G. Withers, P. R.
Rablen, J. A. Hanover, and S. Knapp, Bioorg. Med. Chem., 2010, 18, 7058.
90 S. S. Kale, S. T. Chavan, S. G. Sabharwal, V. G. Puranik and G. J. Sanjayan,
Org. Biomol. Chem., 2011, 9, 7300.
91 S. Mio, R. Ichinose, K. Goto, S. Sugai and S. Sato, Tetrahedron, 1991, 47, 2111.
92 C. J. F. Bichard, E. P. Mitchell, M. R. Wormald, K. A. Watson, L. N.
Johnson, S. E. Zographos, D. D. Koutra, N. G. Oikonomakos and G. W. J.
Fleet, Tetrahedron Lett., 1995, 36, 2145.
93 E. Ösz, L. Somsák, L. Szilágyi, L. Kovács, T. Docsa, B. Tóth and P. Gergely,
Bioorg. Med. Chem. Lett., 1999, 9, 1385.
94 L. Somsák, V. Nagy, T. Docsa, B. Tóth and P. Gergely, Tetrahedron:
Asymmetry, 2000, 11, 405.
95 T. Docsa, K. Czifrák, C. Hüse, L. Somsák and P. Gergely, Mol. Med. Report,
2011, 4, 477.
96 L. Somsák, C.R. Chimie, 2011, 14, 211.
97 L. Somsák, V. Nagy, S. Vidal, K. Czifrák, E. Berzsényi and J.-P. Praly, Bioorg.
Med. Chem. Lett., 2008, 18, 5680.
98 V. Nagy, M. Benltifa, S. Vidal, E. Berzsényi, C. Teilhet, K. Czifrák, G. Batta,
T. Docsa, P. Gergely, L. Somsák and J.-P. Praly, Bioorg. Med. Chem., 2009,
17, 5696.
99 M. Benltifa, J. M. Hayes, S. Vidal, D. Gueyrard, P. G. Goekjian, J.-P. Praly,
G. Kizilis, C. Tiraidis, K.-M. Alexacou, E. D. Chrysina, S. E. Zographos,
D. D. Leonidas, G. Archontis and N. G. Oikonomakos, Bioorg. Med. Chem.,
2009, 17, 7368.
100 T. Tite, L. Tomas, T. Docsa, P. Gergely, J. Kovensky, D. Gueyrard and
A. Wadouachi, Tetrahedron Lett., 2012, 53, 959.
101 M. Benltifa, M. De Kiss, M. I. Garcia-Moreno, C. Ortiz Mellet, D. Gueyrard
and A. Wadouachi, Tetrahedron, Asymmetry, 2009, 20, 1817.
102 C. Laroche, J.-P. Behr, J. Szymoniak, P. Bertus, C. Schütz, P. Vogel and
R. Plantier-Royon, Bioorg. Med. Chem., 2006, 14, 4047.
103 M. S. M. Pearson, R. Plantier-Royon, J. Szymoniak, P. Bertus and J.-B. Behr,
Synhtesis, 2007, 3589.
104 M. S. M. Pearson, N. Floquet, C. Bello, P. Vogel, R. Plantier-Royon, J.
Szymoniak, P. Bertus and J.-B. Behr, Bioorg. Med. Chem., 2009, 17, 8020.
105 S. Mohan and B. M. Pinto, Carbohydr. Res., 2007, 342, 1551.
106 M. Yoshikawa, T. Murakami, H. Shimada, H. Matsuda, J. Yamahara, G.
Tanabe and O. Muraoka, Tetrahedron Lett., 1997, 48, 8367.
107 M. Yoshikawa, T. Murakami, K. Yashiro and H. Matsuda, Chem. Pharm.
Bull., 1998, 46, 1339.
108 K. Jayakanthan, S. Mohan and B. M. Pinto, J. Am. Chem. Soc., 2009, 131,
5621.
109 Y. Li, T.H.-W. Huang and J. Yamahara, Life Sci., 2008, 82, 1045.
110 W. Chen, L. Sim, D. R. Rose and B. M. Pinto, Carbohydr. Res., 2007, 342, 1661.
111 R. Nasi, L. Sim, D. R. Rose and B. M. Pinto, J. Org. Chem., 2007, 72, 180.
112 H. Liu, R. Nasi, K. Jayakanthan, L. Sim, H. Heipel, D. R. Rose and B. M.
Pinto, J. Org. Chem., 2007, 72, 6562.
113 S. Mohanand and B. M. Pinto, Nat. Prod. Rep., 2010, 27, 481.
114 R. Nasi, B. O. Patrick, L. Sim, D. R. Rose and B. M. Pinto, J. Org. Chem.,
2008, 73, 6172.
115 R. Nasi, L. Sim, D. R. Rose and B. M. Pinto, Carbohydr. Res., 2007, 342,
1888.
116 N. S. Kumar, D. A. Kuntz, X. Wen, B. M. Pinto and D. R. Rose, Proteins,
2008, 71, 1484.
260 | Carbohydr. Chem., 2012, 38, 215–262

117 S. Ogawa and M. Kanto, Curr. Top. Med. Chem., 2009, 9, 58.
118 A. Baran and M. Balci, J. Org. Chem, 2009, 74, 88.
119 S. Cantekin, A. Baran, R. Çalis- kan and M. Balci, Carbohydr. Res., 2009, 344, 426.
120 A. Baran, A. Günel and M. Balci, J. Org. Chem, 2008, 73, 4370.
121 R. Lysek, S. Favre and P. Vogel, Tetrahedron, 2007, 63, 6558.
122 A. Scafdi, K. A. Stubbs, R. J. Dennis, E. J. Taylor, G. J. Davies, D. J. Vocadlo
and R. V. Stick, Org. Biomol. Chem., 2007, 5, 3013.
123 A. Trapero, I. Alfonso, T. D. Butters and A. Llebaria, J. Am. Chem. Soc, 2011,
133, 5474.
124 Y-W. Kim, H.-M. Chen, J. H. Kim, J. Müllegger, D. Mahuran and S. G.
Withers, ChemBioChem, 2007, 8, 1495.
125 R. Zhang, J. D. McCarter, C. Braun, W. Yeung, G. D. Brayer and S. G.
Withers, J. Org. Chem., 2008, 73, 3070.
126 K. Slámová, P. Marhol, K. Bezouška, L. Lindkvist, S. G. Hansen, V. Křen
and H. H. Jensen, Bioorg. Med. Chem. Lett., 2010, 20, 4263.
127 S. Dedola, D. L. Hughes, S. A. Nepogodiev, M. Rejzek and R. A. Field,
Carbohydr. Res., 2010, 345, 1123.
128 D. R. da Rocha, W. C. Santos, E. S. Lima and V. F. Ferreira, Carbohydr. Res.,
2012, 350, 14.
129 S. B. Ferreira, A.C.R. Sodero, M.F.C. Cardoso, E. S. Lima, C. R. Kaiser, F.
P. Silva and V. F. Ferreira, J. Med. Chem., 2010, 53, 2364.
130 G. Horne, F. X. Wilson, J. Tinsley, D. H. Williams and R. Storer, Drug
Discov. Today, 2011, 16, 107.
131 T. Kajimoto and M. Node, Curr. Top. Med. Chem., 2009, 9, 13.
132 B. G. Winchester, Tetrahedron: Asymmetry, 2009, 20, 645.
133 T. D. Butters, Curr. Opin. Chem. Biol., 2007, 11, 412.
134 S. D. Boomkamp, J. S. S. Rountree, D. C. A. Neville, R. A. Dwek, G. W. J.
Fleet and T. D. Butters, Glycoconjugate J., 2010, 27, 297.
135 T. Cox T, R. Lachmann, C. Hollak, J. Aerts, S. van Weely, M. Hrebı́cek, F.
Platt F, T. Butters, R. Dwek, C. Moyses, I. Gow, D. Elstein and A. Zimran,
Lancet, 2000, 355, 1481.
136 Z. Luan, k. Higaki, M. Aguilar-Moncayo, L. Li, H. Ninomiya, E. Nanba,
K. Ohno, M. I. Garcı́a-Moreno, C. Ortiz Mellet, J. M. Garcı́a Fernández, and
Y. Suzuki, ChemBioChem 2010, 11, 2453.
137 T. S. Rasmussen, S. Allman, G. Twigg, T. D. Butters and H. H. Jensen,
Bioorg. Med. Chem. Lett., 2011, 21, 1519.
138 G.-N. Wang, G. Reinkensmeier, S.-W. Zhang, J. Zhou, L.-R. Zhang, L.-H.
Zhang, T. D. Butters and X.-S. Ye, J. Med. Chem., 2009, 52, 3146.
139 S. D. Orwing, Y. L. Tan, N. P. Grimster, Z. Yu, E. T. Powers, J. W. Kelly and
R. L. Lieberman, Biochemistry, 2011, 50, 10647.
140 R. F. G. Fröhlich, R. H. Furneaux, D. J. Mahuran, B. A. Rigat, A. E. Stütz,
M. B. Tropak, J. Wicki, S. G. Withers and T. M. Wrodnigg, Carbohydr. Res.,
2010, 345, 1371.
141 R. Khanna, R. Soska, Y. Lun, F. Feng, M. Frascella, B. Young, N. Brignol, L.
Pellegrino, S. A. Sitaraman, R. J. Desnick, E. R. Benjamin, D. J. Lockhart and
K. J. Valenzano, Mol. Ther., 2010, 18, 23.
142 C. A. Denny, K. A. Heinecke, Y. P. Kim, R. C. Baek, K. S. Loh, T. D. Butters,
R. T. Bronson, F. M. Platt and T. N. Seyfried, J. Neurochem., 2010, 113, 1525.
143 J.-P. Praly and S. Vidal, Mini-Rev. Med. Chem., 2010, 10, 1102.
144 T. M. Cox, F. M. Platt and J. M. F. G. Aerts, in Iminosugars-from Synthesis to
Thereapeutic Applications, Ed. P. Compain and O. R. Martin, John Wiley &
Sons, 2007, 295.
145 R. H. Unger, Endocrinology, 2003, 144, 5159.
Carbohydr. Chem., 2012, 38, 215–262 | 261

146 J. M. Aerts, R. Ottenhoff, A. S. Powlson, A. Grefhorst, M. Van Eijk, P. F.
Dubbelhuis, J. Aten, F. Kuipers, M. J. Serlie, T. Wennekes, J. K. Sehti, S.
O’Rahilly and H. S. Overkleeft, Diabetes, 2007, 56, 1341.
147 T. Webbejesm, A. J. Meijer, A. K. Groen, R. G. Boot, J. E. Groener, M. Van
Eijk, R. Ottenhoff, N. Bijl, K. Ghauharali, H. Song, T. J. O’shea, H. Liu, N.
Yew, D. Copeland, R. J. Vn den Berg, G. A. van der Marel, H. S. Overkleeft
and J. M. Aerts, J. Med. Chem., 2010, 53, 689.
148 M. Langeveld, S. A. A. van den Berg, N. Bijl, S. Bijland, C. P. Van Roomen, J.
H. Houben-Weerts, R. Otterhoff, S. M. Houten, K. W. Van Dijk, J. A. Romijn,
A. K. Groen, J. M. Aerts and P. J. Voshol, Metab. Clin. Exp., 2012, 61, 99.
149 N. Bijl, M. Sokolović, C. Vrins, M. Langeveld, P. D. Moerland, R. Ottenhoff,
C. P. A. A. van Roomen, N. Claessen, R. G. Boot, J. Aten, A. K. Groen, J. M.
F. G. Aerts and M. Van Eijk, Hepatology, 2009, 50, 1431.
150 F. Bietrix, E. Lombardo, C. P. A. A. van Roomen, R. Ottenhoff, M. Vos, P. C.
N. Rensen, A. J. Verhoeven, J. M. Aerts and A. K. Groen, Arterioscler.
Thromb. Vasc. Biol., 2010, 30, 931.
151 J. Chang, L. Wang, D. Ma, X. Qu, H. Guo, X. Xu, P. M. Mason, N. Bourne,
R. Moriarty, B. Gu, J.-T. Guo and T. M. Block, Antimicrob. Agents Che-
mother., 2009, 53, 1501.
152 D. Rejman, A. Rabatinová, A. R. Pombinho, S. Kovačková, R. Pohl, E.
Zbornı́ková, M. Kolář, K. Bogdanová, O. Nyč, H. Šanderová, T. Látal, P.
Bartůněk and L. Krásný, J. Med. Chem., 2011, 54, 7884.
153 S. Hakomori, Cancer Res., 1985, 45, 2405.
154 T. M. Wrodnigg, A. J. Steiner and B. J. Ueberbacher, Anticancer Agents Med.,
Chem., 2008, 8, 77.
155 L. Zhang, F. Sun, Y. Li, X. Sun, X. Liu, Y. Huang, L.-H. Zhang, X.-S. Ye and
J. Xiao, ChemMedChem, 2007, 2, 1594.
156 B. Macchi, A. Minutolo, S. Grelli, F. Cardona, F. M. Cordero, A. Mastino
and A. Brandi, Glycobiology, 2010, 20, 500.
157 a) M. Padró, J. A. Castillo, L. Gómez, J. Joglar, P Clapés and C. De Bolós,
Glyconj. J., 2010, 27, 277. b) C. Bello, M. Cea, G.D. Bello, A. Garuti, I. Rocco,
G. Cirmena, E. Moran, A. Nahimana, M.A. Duchosal, F. Fruscione, P.
Pronzato, F. Grossi, F. Patrone, A. Ballestrero, M. Dupuis, B. Sordat, A.
Nencioni, and P. Vogel, Bioorg. Med. Chem., 2010, 18, 3320.
158 R.-J. Wang, C. H. Yang and M.-L. Hu, J. Agric. Food Chem., 2010, 58, 8988.
159 S. A. Yuzwa, M. S. Macauley, J. E. Heinonen, X. Shan, R. J. Dennis, Y. He,
G. E. Witworth, K. A. Stubbs, E. J. McEachern, G. J. Davies and D. J.
Vocadlo, Nature Chem. Biol., 2008, 4, 483.
160 G.-N. Wang, Y. Xiong, J. Ye, L.-H. Zhang and X.-S. Ye, ACS Med. Chem.
Lett., 2011, 2, 682.
262 | Carbohydr. Chem., 2012, 38, 215–262

An overview of key routes for the
transformation of sugars into carbasugars
and related compounds
Raquel G. Soengas,*a José M. Otero,b Amalia M. Estévez,b
Amélia P. Rauter,*c Vasco Cachatra,c Juan C. Estévezb
and Ramón J. Estévez*b
DOI: 10.1039/9781849734769-00263
This chapter reports an overview of the most relevant routes developed for
the transformation of sugars into carbasugars and related compounds.
1 Introduction
Carbasugars, a family of carbohydrate mimics, have attracted great interest
in recent years among chemists and biochemists due to their potential
biological properties.1 The ability of carbasugars to mimic the size, structure
and polarity of parent sugars and their stability towards hydrolysis makes
them potential candidates for the competitive inhibition of glycosidases and
glycosyltransferases.
Most of the synthetical strategies developed for the synthesis of carba-
sugars make use of carbohydrates as starting materials to maintain their
high grade of functionalisation in the final product. It is clear that the use of
carbohydrates provides important advantages for the preparation of their
carbocyclic analogues, since the hydroxyl groups can be maintained
throughout the synthetic sequence, with no need for ‘‘hydroxylation’’
reactions whereas the enantiomeric purity of the target carbasugars will be
guaranteed.
1.1 Carbasugars and pseudo-carbasugars

Carbasugars are carbohydrate analogues in which the endocyclic oxygen
has been replaced by a methylene group (Fig. 1). Natural and synthetic
carbasugars, either as single molecules or as subunits of more complex
molecules, have shown to display interesting biological activities, mainly as
enzymatic inhibitors.2
The carbasugars were introduced about forty years ago when McCasland
and coworkers first synthesised the talopyranose analogue 1,3 followed by
two new isomeric compounds 2 and 34 (Fig. 2).
McCasland used the term ‘‘pseudosugar’’ to designate such analogues
but, as time went by, this term started to be employed to designate a large
variety of sugar analogues. S. Ogawa proposed the use of the prefix
‘‘carba’’, preceded, where considered necessary, by the appropriate locant
a
Department of Chemistry, University of Aveiro, Portugal.
b
Center for Research in Biological Chemistry and Material Sciences (CIQUS), University of
Santiago de Compostela, Spain. E-mail: ramon.estevez@usc.es
c
Carbohydrate Chemistry Group, Center of Chemistry and Biochemistry, Department of
Chemistry and Biochemistry, Faculty of Sciences, University of Lisbon, Portugal.
E-mail: aprauterfcul@gmail.com
Carbohydr. Chem., 2012, 38, 263–302 | 263

O OH OH
HO HO
HO OH HO OH
OH OH
Sugar Carbasugar
Fig. 1 Structures of a six-membered ring sugar and the corresponding carbasugar.
OH OH OH
HO HO HO
HO OH HO OH HO OH
OH OH OH
1 2 3
5a-carba-α-D-talopyranose 5a-carba-α-D-galactopyranose 5a-carba-α-D-gulopyranose
Fig. 2 First three carbasugars synthesised by McCasland’s group.
(‘‘4a’’ for an aldofuranose, ‘‘5a’’ for an aldopyranose), followed by the

name of the sugar.5
However, this definition is sometimes misused by many authors, who
include carbocyclic polyols containing a non-substituted carbon, which
should be called pseudo-carbasugars because they do not have an associated
homomorphic sugar associated.6
1.2 Biological significance of carbasugars and pseudo-carbasugars

Carbasugars are not very abundant in Nature. In fact, the 5a-carba-a-D-
galactopyranose (2), a weak antibiotic against Klebsiella pneumonia MB-
1264, is the only carbasugar isolated from the fermentation broth of
Streptomyces sp. MA-4145.7 However, a large number of carbasugar related
highly oxygenated cyclohexanes of biological interest are present in Nature.
Representative examples include inositols8 (components of the cell-mem-
brane and second messengers); glycosidase inhibitors such as conduritols9
and cyclophellitols;10 aminocarbasugars such as valienamine 7,11 valida-
mine 8, hydroxyvalidamine 9 and valiolamine 10,12 which are components
of natural aminoglycoside antibiotics,13 and ()-shikimic acid (11)14 and
(þ)-quinic acid (12),15 two important biochemical intermediates in plant
and microorganism metabolism (Fig. 3).
Some of these compounds have relevant biological activities and potential
pharmacological applications. For example, (þ)–cyclophellitol 6, isolated
from the culture filtrate of a mushroom strain, Phellinus sp.16 is a potent
inhibitor of b-glucosidases, with potential inhibition of the human immu-
nodeficiency virus (HIV) and with possible antimetastatic therapeutic
activity.17 The (þ)-MK7607 13, isolated from the fermentation broth of
Curvularia eragrostidis D2452,18 is a patented naturally occurring unsatu-
rated pseudo-carbasugar possessing herbicidal activity, and ()-shikimic
acid 11 is a key intermediate in the biosynthesis of amino acids, which has
some pharmacological relevance.19 It has been used as staring material in
the synthesis of many drugs, the most relevant being the antiviral agent
oseltamivir (Tamiflu).20
264 | Carbohydr. Chem., 2012, 38, 263–302

OH OH OH
HO OH OH HO CH2OH
HO OH OH HO
OH OH O
4a 5a 6
myo-Inositol Conduritol D (+)-Cyclophellitol
OH OH OH HO
OH
HO HO HO OH HO
HO NH2 HO NH2 HO NH2 HO NH2

OH OH OH OH
7 8 9 10
Valienamine Validamine Hydroxyvalidamine Valiolamine
CO2H OH
HO CO2H
HO
HO OH HO OH
HO OH
OH OH
OH
11 12 13
Shikimic acid D-(-)-Quinic acid (+)-MK7067
Fig. 3 Examples of naturally occurring pseudo-carbasugars.
Probably, the most important carbasugar derivatives from a biological

point of view are the aminocarbasugars, some of which have become
clinically successful to combat diseases in humans and in plants. In 1970,
researchers at Takeda Chemical Company discovered the agricultural
antibiotic Validamycin A,21 a representative member of the family of
Validamycin antibiotics. The core of these antibiotics is composed by
valienamine (7), an amino pseudo-carbasugar. The pronounced biological
activity of valienamine (7), as well as validamine (8) and hydroxyvalidamine
(9) as glycosidase inhibitors prompted the scientists to synthesize and
evaluate several derivatives.22 Among them are carbocyclic analogues of
glycosylamides, structurally related to glycosphingolipids and glycoglycer-
olipids, which showed similar potencies as immunomodulators, comparable
to those of the corresponding true sugar analogues.23
Carbasugar derivatives are also relevant as potential entities for the elu-
cidation and/or control of biological events involving sugar moieties. The
construction of enzyme substrate analogues, in which part of their structure
has been replaced by a carbasugar unit, is expected to be useful in the
establishement of the mode of action of sugar transferases. An area of
recent interest is the design of potential substrates and inhibitors of fucosyl
and sialyl transferases involved in the last steps of the biosynthesis of Lewis
oligosaccharide antigens.24
2 Strategies for the synthesis of carbasugars and pseudo-carbasugars

The biological significance of carbasugars has stimulated a great interest in
their preparation.1,25 Several approaches have been developed for the
Carbohydr. Chem., 2012, 38, 263–302 | 265

synthesis of carbasugars and related pseudo-carbasugars, the most common
being those that make use of carbanionic and radical cyclizations, ring
closing metathesis (RCM), pinacol couplings, cycloaddition reactions and
rearrangement of cyclic sugars.
2.1 Carbanion cyclisations

The use of anionic species as nucleophiles in the formation of C-C bonds is
widespread in organic chemistry and it has also been extensively applied to
the formation of carbasugars. For this purpose, various stabilising groups
have gained particular attention, namely nitro groups, phosphorous ylides
and aldol precursors.
2.1.1 Malonates. The use of malonate anions to synthesise carbocycles
from open-chain sugars was first introduced by Suami et al.26 Starting from
L-arabinose diethyl dithioacetal tosylate 14, the key iodo-aldehyde 15 was
easily obtained. Reaction of 15 with dimethyl malonate anion, followed by
cyclisation and acetylation provided the cyclohexyl derivative 16 and pyr-
anose derivative 17 in ratio 5:3. Formation of compound 17 could be
explained in terms of a displacement of the iodide by an alkoxide. A
Krapcho decarboxylation of 16, followed by elimination of an acetoxy
group gave the cyclohexene derivative 18. Reduction of the unsaturated
ester with LiAlH4, followed by hydroboration-oxidation and final depro-
tection gave a mixture of 5a-carba-a-D-glucopyranose (19) and 5a-carba-b-
L-altropyranose (20) (Scheme 1).
In order to avoid the competition between addition to the carbonyl and
substitution of the iodide, an alternative pathway incorporating the malonate
moiety prior to the transformation of the primary hydroxygroup into a
leaving group was developed.27 This alternative was used, for instance, for the
preparation of 5a-carba-b-L-mannopyranose pentaacetate (25) (Scheme 2).
2.1.2 Nitro groups. A nitro group is particularly effective at stabilising
carbanions on the adjacent carbon. The nitronate anions, resulting from the
MeO2C CO2Me CO2Me

OBn OBn OBn OBn OAc O
i ii CO2Me
SEt +
CHO
BnO OBn BnO OBn
OTs OBn SEt I OBn
OBn OBn
14 15 16 17
iii
CO2Me
OH OH
HO HO
iv
+
HO OH HO OH
BnO OBn
OH OH
OBn
20 19 18
Scheme 1 Reagents and conditions: (i) a) HgCl2, CaCO3; b) NaI, acetone, reflux, 54%; (ii) a)
dimethyl malonate, NaH, DMF; b) Ac2O, Py, 43% of 16 and 33% of 17 over two steps; (iii)
DMSO, NaCl, heat, 75%; (iv) a) LiAlH4, 15 1C, 83%; b) B2H6, then H2O2, NaOH; c) Ac2O, py,
69% (two steps); d) Na, liq. NH3; e) Ac2O, py; f) NaOMe, MeOH, 65% of 19 and 86% of 20.
266 | Carbohydr. Chem., 2012, 38, 263–302

MeO2C CO2Me
OBn OBn OBn OBn CO2Me AcO
i ii
TBDPSO HO
CHO CO2Me
BnO OBn
OBn OBn
OBn
21 22 23
iii
CO2Me
OAc
AcO
iv
AcO OAc
BnO OBn
OAc
OBn
25 24
Scheme 2 Reagents and conditions: (i) a) dimethyl malonate, py, Ac2O, 85%; b) H2, Raney Ni;
c) TBAF, 48% (two steps); (ii) a) PCC, CH2Cl2; b) Ac2O, Py, 69% (two steps); (iii) DMSO,
NaCl, heat, 70%; (iv) a) LiAlH4, 15 1C, 88%; b) B2H6, then H2O2, NaOH; c) Ac2O, py, 66%
(two steps); d) Pd-black, MeOH; e) Ac2O, py, 57% (two steps).
abstraction of the proton in the a position (pKa 9-10) are bidentate

nucleophiles stabilised from resonance. On account of this nucleophilicity,
nitroderivatives are highly valuable intermediates for the formation of C-C
bonds and have been widely used in carbohydrate chemistry.28 The most
relevant C-C bond formation reactions promoted by the nucleophilic
addition of nitronates reactions are the Henry reaction (addition to an
aldehyde or ketone),29 the C-alkylation (addition to a saturated electrophilic
carbon)30 and the Michael addition (addition to an a,b-unsaturated sys-
tem).31 In particular, reactions involving cyclisation of nitrosugars are useful
processes in the enantioselective preparation of carbasugars.32
In 1984 Kitagawa and co-workers showed that an intramolecular Henry
reaction of a nitrosugar is a useful process for the enantioselective
preparation of carbasugars (Scheme 3).33 Thus, addition of nitromethane to
sugar derivative 26 afforded epimeric alcohols 27 and 28. Removal of
hydroxyl group of 27 by reductive deacetoxylation and subsequent removal
of the acetal protecting group provided nitro sugar 29, which in basic
conditions cyclised to a pentahydroxylated nitro cylohexane, which was
then easily converted into derivative 30. Finally, radical-mediated denitra-
tion and subsequent removal of the di-O-isopropylidene groups gave
5a-carba-a-D-glucopyranose (31). On the other hand, nitro sugar 28 was
similarly converted into 5a-carba-b-L-idopyranose (20), via compounds 32
and 33.
Since then, this nitro sugar based protocol has been widely used for the
preparation of carbasugar and related compounds.34
A recent example is the synthesis of hydroxymethyl-myo-inositol 37,
which was easily obtanided from a D-glucose-derived aldehyde 34. Firstly
fluoride catalysed Henry reaction of nitroethanol and aldehyde 34 afforded
nitro sugar epimers 35a. Benzylation of the free hydroxyl groups afforded
the nitrosugar mixture 35b, which, after liberation of their anomeric posi-
tions by acidic hydrolysis, were submited to an intramolecular Henry
reaction using hydrogen sodium carbonate as a base. The result was the
Carbohydr. Chem., 2012, 38, 263–302 | 267

BzO BzO BzO
O i HO O HO O
O O O + O
O2N O2N
O BnO O BnO O
BnO
26 27 28
ii ii
BzO BzO
O O
OH OH
O2N O2N
BnO OH BnO OH
29 32
iii iii
NO2 NO2
OH O O O O HO OH
HO iv iv
OH O O O O HO OH
HO
OH OBn OBn OH
31 30 33 20
Scheme 3 Reagents and conditions: (i) CH3NO2, NaH, 15-crown-5, DMF, 61%; (ii) a) Ac2O,
TsOH, NaBH4, EtOH, 69%; b) 80% aq AcOH, 80 1C, 93% of 29 and 42% of 32; (iii) a) KF,
18-crown-6, DMF then NaOMe, MeOH; b) 2,2-dimethoxypropane, TsOH, CuSO4, 72% of 30
and 42% of 33; (iv) a) n-Bu3SnH, AIBN, Benzene, 80 1C; b) aq AcOH; c) Na, liq NH3, 30% of
31 and 20% of 34.
O O2N O2N
OH OBn
O i O ii O
H O O O
HO BnO
BnO O O O
BnO BnO
34 35a 35b
iii
CH2OBn O2N CH2OBn O2N CH2OBn

BnO OH BnO OEE BnO OH
v iv
HO OH EEO OEE HO OH
OBn OBn OBn
37 36b 36a
Scheme 4 Reagents and conditions: (i) HO(CH2)2NO2, TBAF, THF; (ii) BTCA, TfOH, C6H12,
CH2Cl2; (iii) a) TFA/dioxane/H2O; b) HNaCO3 2% MeOH: (iv) CH2¼CHOCH2CH3, C6H12,
CH2Cl2; (v) a) HSnBu3, AIBN, toluene, reflux; b) PPTS, EtOH, 50 1C.
specific formation of nitrocyclohexane 36a, which was finally transformed

into desired inositol 37 by removal of the nitro group (Scheme 4).35
Recently, El Blidi et al. have publised the synthesis of a nitrocyclohexane
derivative by means of a chemoenzymatic variant of the classical method of
inter-intramolecular Henry reaction (Scheme 5).36 Nitroaldol cyclisation
between a masked 3-hydroxy-4-nitrobutyraldehyde 38 and dihydrox-
yacetone phosphate (DHAP) 39 catalysed by fructose-1,6-biphosphate
aldolase (RAMA) produced the nitrocyclohexane 40.
Another useful strategy for the preparation of carbasugars from nitro
sugars involves a Michael addition-Henry reaction sequence.
268 | Carbohydr. Chem., 2012, 38, 263–302

OH
OEt OH
O i
O2N HO NO2
+ HO OPO3
OH OEt
HO OH
38 39
40
Scheme 5 Reagents and conditions: (i) Fructose-1,6-biphosphate aldolase.
CO2 Et
EtO2 C EtO2 C
O i O ii HO NO2
O2 N OH
O
O BnO OH
BnO BnO OH
41 42 43 OH
iii
HOH 2C CH 2OH
HO
i, iii, ii O i, ii
NO 2 O2 N O HO NO2
BnO
BnO O BnO OH
HO OH
45 44 46 OH
Scheme 6 Reagents and conditions: (i) CH3NO2, K2CO3, EtOH; ii) TFA/H2O; (iii) DABCO,
benzene, reflux; (iv) LAH, THF; (v) NaIO4.
For example, a Michael addition of nitromethane to a,b-unsaturated

ester 41 yielded nitroester 42 (Scheme 6). An intramolecular nitroaldol
reaction of compound 42 with DABCO in refluxing benzene afforded
nitrocyclohexane 43 in high yield, as the only diastereomer. This remarkable
diastereoselectivity was also observed in the transformation of the sugar
derivative 44 into nitrocyclopentane 45 and nitrocyclohexane 46, when
proceeding as stated in Scheme 6.37
On the other hand, nitro olefins are powerful electrophilic species that
react with nucleophiles to give 1,2-addition products, which can then react
intramolecularly with aldehyde groups to render carbocycles. This strategy
was recently used for the preparation of carbasugar rancinamycin (52).38
Michael addition of 1,3-dithiane to sugar nitro olefin 47 provided a separ-
able mixture of epimers 48 and 49. The major isomer 49 was subjected to a
reaction protocol including an intramolecular Henry reaction to give a
mixture of epimeric carbocycles 50a and 51, which when reacted with 2,2-
dimethoxypropane and PTSA produced only its derivative 50b only.
Finally, treatment of 50b with MeI and aqueous sodium hydrogen carbo-
nate gave compound 52, as a result of the removal of the carbonyl pro-
tecting group followed by a kinetically and thermodynamically favoured
E1CB elimination of the nitro group (Scheme 7).
C-Alkylation of nitronates has also been applied to the transformation of
nitro sugars into carbasugars. Although O-alkylation is usually favoured,
C-alkylation is the predominant process when it is the thermodynamically
Carbohydr. Chem., 2012, 38, 263–302 | 269

S O2N O2N
O2N Li
S O O
O DTN O + DTN O
O i
BnO O BnO O
BnO O
47 48 49
ii, iii
OBn OBn OBn

HO HO OH HO OH
O iv
+
DTN O OH OH
DTN DTN
NO2 NO2 NO2
50b 50a 51
OBn
HO
O
OHC O
52
Scheme 7 Reagents and conditions: (i) THF, 78 1C; (ii) 75% AcOH, reflux; (iii) 2% aq.
HNaCO3, MeOH; (iv) (CH2)3C(OMe)2, p-TsOH, acetone; (v) MeI, aq. HNaCO3, MeCN, 35 1C.
O2N O2N OBn

O i O ii
O OMe
BnO BnO O NO2
O MeO OBn
BnO BnO OTf
53a 53b 54
iii
OH OBn OBn
HO v iv
OH
O NHZ O NHZ
NH2
HO HO OBn MeO OBn
56 55b 55a
Scheme 8 Reagents and conditions: (i) a) AcCl/MeOH; b) Tf2O, py, CH2Cl2; (ii) TBAF, THF;
(iii) a) H2, Raney Ni, MeOH; b) NaCO3H, ClCO2Bn, AcOEt; (iv) TFA, H2O; (v) a) NaBH4,
MeOH; b) H2, Pd/C, AcOH.
favoured process. Thus, as stated in Scheme 8 similar to the synthesis of

aminocyclopentitol shown in Scheme 6 an stereocontrolled intramolecular
alkylation of the nitronate of the sugar trifil derivative 53b, provided the
novel bicyclic compound 54. Catalytic hydrogenation of 54 afforded the
corresponding amino acetal, which was directly reacted with benzyl chlor-
oformate and sodium bicarbonate to obtain its derivative 55a. Subsequent
hydrolysis of this bicyclic glycoside with trifluoroacetic acid was followed by
the immediate reduction using NaBH4 of the implicit carbonyl group of the
270 | Carbohydr. Chem., 2012, 38, 263–302

resulting bicyclic hemiacetal 55b and final deprotection, resulting in the
amino carbasugar 56 (Scheme 8).39
This synthetic methodology was also applied to the preparation of several
polihydroxylated cyclopenthyl b-amino acids.40
2.1.3 Reactions of phosphorus-stabilised species. The cyclisation of car-
banions stabilised by phosphonate neighbouring groups have been exten-
sively used in the synthesis of carbapyranoses from carbohydrates. Special
attention was devoted to the intramolecular Horner-Wadsworth-Emmons
olefination (HWE)41 since this strategy for the preparation of sugar-derived
carbocycles was introduced by Paulsen et al. (Scheme 9).41a Reaction of the
glucose-derived aldehyde 57 with lithium dimethyl methyl phosphonate and
further manipulation of protecting groups afforded the intermediate 58a.
Swern oxidation followed by spontaneous cyclisation yielded enone 59a,
which upon treatment with sodium borohydride produced the alcohol 60a
as the major product. O-Desilylation, catalytic hydrogenation, O-deben-
zylation and acetylation converted 60a into 5a-carba-b-D-glucopyranose
pentaacetate (61) and 5a-carba-a-L-idopyranose pentaacetate (62).
A HWE olefination was the key step in a recent synthesis of gabosine I
from D-glucuronolactone described by Shin et al.42 D-Glucuronolactone was
differentially protected as a trans-diacetal at C-2,3, an ethoxymethyl ether at
C-4 and a tert-butyldimethylsilyl ether at C-6, to give its derivative 63
(Scheme 10). A nucleophilic addition of lithiated dimethyl methyl phos-
phonate to the lactone carbonyl afforded phosphonate-lactol 64 in 96%
yield. The hemiacetal was reduced by sodium borohydride and the resulting
derivative 58b was subjected to a Swern oxidation to produce the diketone
66. An intramolecular HWE olefination was effected in the same pot to give
enone 59b in 78% overall yield. Complete deprotection of 59b left gabosine
I (59c).
2.1.4 Aldol cyclisations. Aldol condensations are important carbon–
carbon bond formation tools in organic synthesis. Tadano et al. introduced
OTBDPS
O O
OH TBDPSO
O i OBn ii
OBn
BnO OBn
BnO PO(OMe)2
BnO CHO OBn
OBn OH
OBn
57 58a 59a
iii
OAc OAc OH
AcO AcO TBDPSO
iv
+
AcO OAc AcO OAc BnO OBn
OAc OAc OBn
62 61 60a
Scheme 9 Reagents and conditions: (i) a) CH3P(O)(OMe)2, n-BuLi, THF, 78 1C. b) AcOH,
60 1C. c) TBDPSCl, imidazole, 85% (three steps). (ii) (COCl)2, DMSO, Et3N, 78 1C, 50%.
(iii) NaBH4, 87%. (iv) a) TBAF, THF. b) H2, Ni-Ra, dioxane. c) H2, Pd/C. d) Ac2O, py, 54%
(four steps).
Carbohydr. Chem., 2012, 38, 263–302 | 271

TBSO TBSO TBSO
O O OH O OH OH O
EOMO O EOMO OMe EOMO
i P ii P OMe
OMe OMe
O O O O O O
MeO OMe MeO OMe MeO OMe
63 64 58b
iii
TBSO TBSO
HO O O O
EOMO O EOMO
v iv P OMe
HO O OMe
O O O O
HO OH
MeO OMe MeO OMe
59c 59b 66
Scheme 10 Reagents and conditions: (i) CH3P(O)(OMe)2, LDA, THF, 78 1C, 96%.
(ii) NaBH4, MeOH, 0 1C, 96%. (iii) a) TFAA, DMSO, CH2Cl2, 78 1C. b) DIPEA, 78 1C.
(iv) TEA, LiCl, r.t., 78%. (v) TFA, H2O, r.t., 96%.
O O O
O i O ii O
O O O O O
O O
O O O
O
67 68 69
iii
OBn H
OAc H
BnO O O
AcO OAc v iv
O O
OAc BnO O
AcO H H O
O
72 71 70
Scheme 11 Reagents and conditions: (i) a) Ph3P¼CHCOCH3, benzene, 96%; b) H2, Raney-Ni,
MeOH; c) PCC, CH2Cl2, 88%; (ii) a) 60% aq. AcOH, 95%; b) NaIO4, H2O, MeOH, quant.;
(iii) DBU, benzene then Ac2O, py, 44%. (iv) a) H2O2, NaOH, MeOH, 96%; b) NaBH4, EtOH,
84%; c) 2-methoxyethanol, H2O, NaOAc; d) NaH, BrBn, DMF; (v) a) AcOH, H2O, dioxane
then NaBH4, EtOH; b) NaIO4, H2O, MeOH; c) Ac2O, py, 44%.
this methodology for the preparation of several carbasugars (Scheme 11).43

Thus, a Wittig olefination of compound 67 with acetylmethylene-
triphenylphosphorane, followed by catalytic hydrogenation and oxidation
provided ketone 68. Hydrolysis of the 5,6-O-isopropylidene moiety, followed
by oxidative cleavage of the diol afforded di-carbonyl compound 69, which on
treatment with DBU followed by acetic anhydride and pyridine afforded the
intramolecular aldol condensation product 70. Epoxidation of 70, reduction
with sodium borohydride and epoxide ring opening with a hydroxide anion
gave, after benzylation of the free alcohols, the diaxial opening product 71.
Finally, glycol cleavage, reduction, removal of the protecting groups, and
acetylation yielded 5a-carba-a-L-altropyranose pentaacetate (72).
272 | Carbohydr. Chem., 2012, 38, 263–302

O OTBS O O
O 74 i O ii O
CHO H H
O O TBSO
O OH TBSO OTBS
73 75 76
iii
O O
OH O
HO v iv O
H H H
HO OH O
HO OTBS
OH
OTBS TBSO OTBS
3 78 77
Scheme 12 Reagents and conditions: (i) BF3 Et2O, CH2Cl2, 80 1C, 75%; (ii) a) H2, Pd/C,
THF, 91%; b) TBSOTf, 2,6-lutidine, CH2Cl2, 90%; c) 70% aq. AcOH, 50 1C, 96%; d) TBSCl,
imidazole, py, 70%; (iii) a) 80% aq. AcOH, 80%; b) (COCl)2, DMSO, Et3N, CH2Cl2, 80 1C,
97%; (iv) LDA, THF, 80 1C, 51%; (v) a) TESOTf, DMAP, py, 95%; b) LiAlH4, THF, 0 1C,
90%; c) 6N HCl/THF/MeOH, 95%.
A synthesis of 5a-carba-a-D-gulopyranose (3) from D-glyceraldehyde

was later reported. As shown in Scheme 12, protected derivative
D-glyceraldehyde 73 was reacted via a vinylogous cross aldolisation with
furan silyloxy diene 74 to selectively give the key alcohol 75. Reduction of the
double bond, followed by silylation of the free hydroxyl, removal of the
isopropylidene protecting group and full protection of the free hydroxyls
with silyl groups, lead to 76. Selective desilylation of the primary hydroxyl,
followed by Swern oxidation afforded the di-carbonyl compound 77, which
under basic conditions was transformed into the bicyclic lactone 78 as a single
stereoisomer. Silylation of the generated alcohol, reduction of the lactone
with LiAlH4 and desilylation, gave 5a-carba-b-D-gulopyranose (3).44
More recently, several gabosines have been synthesised from D-glucose
via an intramolecular direct aldol reaction of a 2,6-diketone.45 Thus, an
intramolecular aldol cyclisation of D-glucose derived diketone 79,46 pro-
vided enone 59d, that was reduced with K-selectride and the resulting
alcohol treated with TBSCl and imidazole to give a silyl ether 80a (Scheme
13). The isopropylidene blocking group in 80a was selectively removed with
80% aqueous acetic acid to afford the corresponding diol, which on
regioselective mesylation followed by treatment with Super hydride
(LiEt3BH) afforded alcohol 81. PDC oxidation of 81 gave enone 82a in a
high yield. (þ)-Gabosine A (82b) was the final result after acidic hydrolysis
of 82a.
Tatsuta and co-workers described a procedure for the cyclisation of
carbohydrates to carbasugar derivatives based on a SnCl4-promoted aldol-
like cyclisation of silylenol ethers.47 Thus, D-xylonolactone derivative 83
was reacted with lithiated methyl phenyl sulfone to give furanose 84
(Scheme 14). Compound 84 was converted into cyclohexenone 85 by ring
opening with TBSOTf and SnCl4-promoted intramolecular reaction. A
Michael reaction with tributylstannyl lithium followed by the trapping of
the produced anion with formaldehyde and subsequent desulfonylation
Carbohydr. Chem., 2012, 38, 263–302 | 273

O O O
O
O i O O ii O OTBS
O
O O O O O O
MeO OMe MeO OMe MeO OMe
79 59d 80a
iii
H3C H3C
H 3C
v O OTBS iv HO OTBS
O OH
O O O O
HO OH
MeO OMe MeO OMe
82b 82a 81
Scheme 13 Reagents and conditions: (i) a) L-Proline, DMSO, 83%; b) POCl3, py, 99%; (ii) a)
K-selectride, THF, 78 1C, 99%; b) TBSCl, imidazole, DMF, 95%; (iii) a) 80%, AcOH, 88%;
b) MsCl, 2,4,6-collidine, CH2Cl2, 78 1C, 92%; c) LiEt3BH, THF, 78 1C, 84%; (iv) PDC,
3A, MS, CH2Cl2, 92%; (v) TFA, H2O, CH2Cl2, 94%.
MeO MeO PhO2S

O O OH
i O ii
MeO SO2Ph
MeO O OTBS
HO OH
TBSO OTBS TBSO OTBS
83 84 85
iii
HO MOMO HO
v iv
HO NH2 MOMO OH O OTBS
HO OH HO OH TBSO OTBS
87 60b 86
Scheme 14 Reagents and conditions: (i) MeSO2Ph, n-BuLi/THF, 78 1C, 0.5 h, 94%; (ii) a)
TBSOTf, 2,6-lutidine/CH2Cl2, 40 1C, 48 h, 92% b) SnCl4/CH2Cl2, 78 1C, 3 h, 70%; (iii) n-
Bu3SnLi, HCHO/THF, 78 1C to 40 1C, 72 h, 84%; (iv) a) Zn(BH4) 2/Ether, 0 1C, 1 h, 80%; b)
MOMCl, n-Bu4NI, DIPEA CH2ClCH2Cl, 50 1C, 24 h, 85%; c) TBAF/THF, rt, 3 h, 97%; (v) a)
HN3, Ph3P, DEAD/THF, rt, 1 h, 81%; b) H2, Raney-Ni, H2O, 1,4-dioxane, quant; c) 3% HCl,
MeOH, 99%.
yielded the a-hydroxymethylcyclohexenone 86. Stereoselective reduction of

the carbonyl group in 86 followed by protection and deprotection steps led
to cyclohexene 60b. Valienamine (87) was produced from 60b by Mitsunobu
inversion of the allyl alcohol and deprotection.47b
An iron-catalysed tandem isomerisation–intramolecular aldolisation of
vinyl pyranoses reaction was developed for the preparation of carbohy-
drate-derived cyclohexenones. Thus, treatment of D-glucose derived vinyl
pyranose 88 with 10 mol% of Fe(CO)5 under irradiation with a HPK 125
274 | Carbohydr. Chem., 2012, 38, 263–302

O Me O Me
O i ii
BnO OH BnO OH BnO
BnO OBn BnO OBn BnO OBn

88 89a 90a
iii
O Me
HO
HO OH
90b
Scheme 15 Reagents and conditions: (i) Fe(CO)5 (10 mol%), THF, hn, 95%; (ii) MsCl, Et3N,
CH2Cl2, 69%; (iii) FeCl3, CH2Cl2, 55%.
HO
OBn OBn S S S BnO S
i or ii HO2C NHTfa
+
S
O BnO S
OBn NHTfa BnO NHTfa
OBn BnO
91 92 93
Scheme 16 Reagents and conditions: (i) n-BuLi, THF, 10 1C; (ii) n-BuLi, THF, 1 eq. LiBr,
90 1C to 50 1C.
lamp afforded cyclohexanone 89a as a mixture of stereoisomers (Scheme

15).48 Intermediate 90a was obtained from 89a through the corresponding
mesylate. Then, removal of the benzyl groups with FeCl3 in CH2Cl2
afforded the desired 4-epi-gabosine A (90b).
2.1.5 Sulfur-stabilised carbanions. The Corey–Seebach method was
used for the intramolecular anionic cyclisation of sugar derivatives,49 as in
the case of the D-glucosamine derivative 91. Treatment with 2.5 eq. of n-
BuLi in THF, resulted in cyclisation to the carba analogue of 1-amino-1-
deoxy-b-L-galacto-pyranose derivative 92, accompanied by a certain
amount of the carba analogue of the 1-amino-1-deoxy-a-D-altro-septanose
derivative 93 (Scheme 16).50 Under standard Corey–Seebach conditions,
compounds 92 and 93 were obtained in 54% and 23% yield, respectively.
However, when LiBr was added, the cyclization was speeded up and the
ratio of the six-membered product increased significantly, probably via a Li-
chelated intermediate (62% yield for compound 92 and 8% yield for
compound 93).
2.2 Carbocation cyclisations

Even though cyclisation processes involving carbanions are more common,
methods making use of carbocations also proved to be successful in the
synthesis of carbasugars. For example, Mootoo et al. reported a synthesis of
b-carba-galacto-disaccharides involving a key step consisting of the
Carbohydr. Chem., 2012, 38, 263–302 | 275

OCHO CO 2H
O OH
i OAc ii SPh
BnO
O O
O
O O
O
94 95 96
iii
OH OBn OBn OBn
v iv O
HO OR O OR O O OMe
HO OH O O SPh BnO OBn

99 98 97
Scheme 17 Reagents and conditions: (i) DIB, I2, cyclohexane, 95%; (ii) a) PhSH, BF3 OEt2,
78 1C, then NaOMe, MeOH; b) NaH, BnBr, TBAI, DMF, 70% (two steps); c) O3, MeOH-
CH2Cl2, 78 1C, then Ph3P; d) NaClO2, CH3CN, 2-methyl-2-butene, 59% two steps; (iii)
DCC, DMAP, methyl 2,3,6-tri-O-benzyl-a-D-glucopyranoside, 80%; (iv) a) Tebbe reagent,
80%; b) MeOTf, DTBMP, CH2Cl2, 64%; (v) a) BH3 SMe2, then H2O2, NaOH, 72%; b) HCl,
MeOH; c) Pd/C, EtOH, HCO2H, 72% (two steps).
formation of the carbasugar ring via an oxycarbenium ion-enol ether

cyclisation.51 Thus, a Suarez fragmentation (Scheme 17) of D-lyxose-derived
branched pyranoside 94 resulted in the formation of lactol 95, which was
transformed into the mixture of acids 96. A DCC-mediated esterification
was employed to introduce the glycon moiety and the key enol ether 97 was
obtained after a Tebbe reaction. Activation with methyl triflate in the
presence of 2,6-di-tert-butyl-4-methylpyridine led to the cyclic enol ether 98.
Finally, treatment of 98 with BH3 SMe2 and deprotection steps provided
the carbadisaccharide 99.
2.3 Radical cyclisations

Radical cyclisations have been extensively used to convert carbohydrates
into carbocycles.52 The most extended strategy consists of opening the
sugar, transforming the carbonyl into a radical acceptor (alkene or alkyne)
and converting one of the hydroxyl groups into a nucleophilic radical. On
other hand, the most common radical generators are tributyl tin hydride
and samarium diiodide.
An early contribution was a comprehensive study on the application of
the radical cyclisation of hept-1-enitols to the preparation of 5a-carbasugars
published by Redlich and co-workers in 1992.53 Starting from D-ribose
derivative 100a, they prepared different substrates and studied the outcome
of the n-Bu3SnH/AIBN promoted radical cyclisation reaction. For example,
radical cyclisation of iodoalkene 102a afforded cyclohexane 103, together
with some deiodination product 102b (Scheme 18).
On the other hand, it is known that the reductive dealkoxyhalogenation
of 6-deoxy-6-iodoglycosides (Grob fragmentation) gives rise to ring-opened
hex-5-enals,54 which are suitable substrates for ketyl-olefin radical cyclisa-
tion. Thus, 5-hexenals obtained by treatment of the corresponding methyl
6-deoxy-6-iodoglycosides with powdered zinc have been employed as
276 | Carbohydr. Chem., 2012, 38, 263–302

OTBDPS OTBDPS
OH BnO I
O OH i ii
O OH O OBn
O O O
O
100a 101 102a
iii
OBn OBn
BnO + BnO
O O
O O
103 102b
Scheme 18 Reagents and conditions: (i) vinylmagnesium bromide, THF, 79%; (ii) a) NaH,
BrBn, DMF, 62%; b) TBAF, THF; c) I2, PPh3, imidazole, 79%; (iii) Bu3SnH, AIBN, benzene.
O OMe CH3 CH3

I i BnO O ii BnO OH + BnO OH
BnO OBn
OBn BnO OBn BnO OBn BnO OBn
104a + 105a 106 107 108
Scheme 19 Reagents and conditions: (i) Zn, EtOH; (ii) SmI2, THF.
I CH3
O i AcO i
O AcO OH
AcO OAc
AcO AcO AcO

109 110 111
Scheme 20 Reagents and conditions: (i) SmI2, THF.
substrates for a SmI2-mediated radical cyclisation, allowing the overall

transformation of pyranoses into cyclopentanes. For example, reaction of
D-glucose derived 5-hexenal 106 with SmI2 under dilute conditions in the
presence of a proton source provided the desired cyclopentanols 107 and
108 in a 2:1 ratio and in good yield (Scheme 19).55
A modified and more efficient one-pot procedure, in which Grob frag-
mentation was executed under the action of samarium diiodide, was suc-
cessfuly applied to the preparation of cyclopentane derivative 111, as
depicted in Scheme 20.56
Similarly, Chiara et al. reported a cascade reaction of 6-deoxy-6-iodo-
hexopyranosides promoted by samarium diiodide.57 A series of 6-deoxy-6-
iodohexopyranosides with different configurations and substitution patterns
were treated with SmI2 in THF/HMPA to give the desired cyclopentanols.
For example, reaction of iodopyranoside 112 with SmI2 in THF and HMPA
afforded cyclopentanols 113 and 114, together with the dehalogenation
product 115 (Scheme 21).
Carbohydr. Chem., 2012, 38, 263–302 | 277

O O
O I Me Me O
i Me
O O O O
O O
TBSO OMe OH OH TBSO OMe
TBSO TBSO
112 113 114 115
47% 17% 11%
Scheme 21 Reagents and conditions: (i) SmI2, HMPA, THF.
O CH2OH
O OMe
H i BnO OH
BnO OBn
OBn BnO OBn
116a 117
Scheme 22 Reagents and conditions: (i) SmI2, HMPA, THF, t-BuOH.
O OH I N NHOMe
i OMe ii
PivO PivO PivO

O O O
O O O
118 119 120
Scheme 23 Reagents and conditions: (i) a) NH2OMe, MeOH; b) I2, PPh3, toluene; (ii) SmI2,
HMPA, THF. 72%.
6-Formyl methyl pyranosides undergo a ring contraction induced by

samarium(II) iodide that results in the formation of highly functionalised
cyclopentanes. Thus, treatment of methyl pyranoside 116a with samar-
ium(II) iodide in the presence of HMPA and tert-butyl alcohol yielded
cyclopentane 117 (Scheme 22).58
An efficient method for the preparation of highly functionalised amino-
cyclopentanes from readily available carbohydrate templates was reported in
1997. In this procedure, a radical generated by treatment of an iodoalkane
and samarium diiodide is trapped intramolecularly by an oxime ether, thus
constructing a cyclopentane ring.59 For example, carbohydrate hemiacetal
118 was condensed with O-methylhydroxylamine to give, after iodination of
the released hydroxyl group, iodo oxime 119. The samarium diiodide-
mediated reductive cyclisation of 119 in a THF solution containing HMPA
furnished the alkoxy aminocyclopentane 120 in good yield (Scheme 23).
A full paper published by Chiara et al. in 1997, describes a series of SmI2-
mediated intramolecular reactions that involve trapping radicals by sugar
derived oxime ethers, comparable with the transformation of oxime deri-
vative 121 into amino cyclopentane 122 (Scheme 24).60
Particularly relevant is the work carried out by Gómez et al., designed to
allow a direct transformation of carbohydrates into the corresponding
carbasugars.61 A representative example depicted in Scheme 25 involves a 6-
exo-dig radical cyclisation involving the alkyne moiety (radical acceptor)
278 | Carbohydr. Chem., 2012, 38, 263–302

Br
i AcO NHOBn
AcO NOBn
O O
O O
121 122
Scheme 24 Reagents and conditions: (i) SmI2, HMPA, THF, 40 1C, 61%.
Ph Ph
OPh
O OH S
O
i OH ii O
O OH O OTBS
O O
O O O O O
O O
123 124a 124b
iii
Ph Ph
OTBS OTBS
O O
+
O O O O
O O
125 126a
Ph OH
OBn OAc
OBn O AcO
iv O v vi
126a
O O AcO OAc
O O
O OAc
O
126b 127 128
Scheme 25 Reagents and conditions: (i) PhCCLi, THF, 78 1C, 65%; (ii) a) TBSCl, Et3N,
DMAP, CH2Cl2, 65%; b) ClC(S)OPh, py, CH3CN, 80%; (iii) n-Bu3SnH, AIBN, toluene, 90 1C,
95%; (iv) a) TBAF, THF, 91%; b) NaH, BrBn, Bu4NI; (v) a) O3, MeOH/CH2Cl2 (1:1), 78 1C,
then Me2S; b) BH3 SMe2, THF, 75% (3 steps); (vi) a) NaH, CS2, MeI; b) n-Bu3SH, AIBN,
toluene, 85% (2 steps); c) H2, Pd/C, MeOH; d) AcOH-THF-H2O (4:2:1), 60 1C; e) Ac2O, py,
85% (3 steps).
and the thionocarbonate moiety (radical precursor) of the D-mannose

derivative 124b.59b This was easily prepared by treatment of D-mannose
diacetonide 123 with lithium phenylacetylide and then converted, in two
steps, into phenyl thionocarbonate 124b. Treatment of 124b with n-Bu3SnH
and AIBN resulted in a radical cyclisation leaving a separable 3:2 mixture of
exo methylenecyclohexanes 125 and 126a. Cyclohexane 126a was trans-
formed into its derivative 126b, which upon ozonolysis followed by reduc-
tion of the resulting ketone lead to cyclohexane 127. Finally, compound 127
was transformed into carbasugar 128 in five steps.
2.4 Pinacol couplings

Pinacol coupling of carbonyl compounds has been extensively studied62
and can be promoted by a wide range of organometallic reagents and
Carbohydr. Chem., 2012, 38, 263–302 | 279

low-valence lanthanide salts such as samarium diiodide, activated manga-
nese, titanium derivatives and zinc powder.63 This ketyl radical cyclisation
seemed to be particularly suitable for the preparation of polyhydroxylated
carbocyles from carbohydrates and specifically those promoted by samar-
ium diiodide, which have been extensively used in carbohydrate chemistry.64
This reducing agent has proven to be particularly useful for promotion of
intramolecular pinacol coupling reactions of polyoxygenated substrates
under very mild conditions, in high yield and with high diastereoselectivity.
2.4.1 Intramolecular pinacol coupling of aldehydes and ketones. After

being reported in 1991 that carbohydrate-derived dialdehydes undergo a
pinacol coupling reaction in the presence of SmI2 to give cis-diols as the
preponderant products,65 this intramolecular coupling reaction was exten-
sively applied to the preparation of polyhydroxylated cycloalkanes and
derivatives of biological interest, such as inositols,66 carbasugars67 and
aminocycloalkanes.68 For example, in 1994 Chiara et al. reported that a
tandem Swern oxidation/SmI2-mediated intramolecular pinacol coupling of
D-mannitol derivative 129 resulted in a 98:2 mixture of inositols 4b and 130
(Scheme 26).66a
A samarium diiodide-mediated intramolecular pinacol coupling of a
carbohydrate derived dialdehyde was also the key step in a route designed
for the preparation of cyclopentane-based congeners of myo-inositol.66c
Treatment of dialdose 131 with an excess of samarium diiodide and t-BuOH
in THF gave a 1:3 mixture of cyclopentitols 132a and 133 (Scheme 27).
The key step of the synthesis of the natural carbasugar caryose, reported
in 1997, was also a samarium diiodide promoted pinacol coupling.67a Thus,
the SmI2-mediated cyclisation of ketoaldehyde 135, derived from 3,4,5-tri-
O-benzyl-l-deoxy-D-iditol 134, followed by a mild oxidation of the cis-l,2-
diol 132b and a diastereoselective allylation of the ketol 136, gave 137, a
precursor of caryose (138) (Scheme 28).
Recently, a samarium diiodide-mediated intramolecular pinacol coupling
of a carbohydrate-derived dialdehyde was also the key step in a route
t
Bu(Ph2)SiO O
D-Mannitol HO
OH
O OSi(Ph2)tBu
129
t t
Bu(Ph2)SiO Bu(Ph2)SiO
O OH O OH
O OH O OH
t
t
Bu(Ph2)SiO Bu(Ph2)SiO
4b (92 : 8) 130
Scheme 26 Reagents and conditions: (i) (COCl)2, DMSO, Et3N, CH2Cl2. (ii) SmI2, t-BuOH,
THF.
280 | Carbohydr. Chem., 2012, 38, 263–302

OPMB OPMB
O
PMBO i PMBO OH PMBO OH
O
PMBO
OBn
BnO OH BnO OH
131 132a 133
Scheme 27 Reagents and conditions: (i) SmI2, t-BuOH, THF, 43%.
OBn
BnO CHO i BnO CHO ii
BnO OH BnO O BnO OH
BnO BnO
CH3 CH3 OH
BnO CH3
134 135 132b
iii
OH OH OBn OBn
HO BnO iv BnO O
OH CHO OH
OH OH OH
HO CH3 BnO CH3 BnO CH3
138 137 136
Scheme 28 Reagents and conditions: (i) (COCl)2, DMSO, Et3N, CH2Cl2; (ii) SmI2, t-BuOH,
THF; (iii) TEMPO, KBr, 0.9M NaOCl, CH2Cl2; (iv) AllylTMS, TiCl4, CH2Cl2.
OBn OBn OBn

BnO i BnO BnO
+
BnO BnO BnO
O OH OH
139 140a 141
ii
OBn OBn OBn

BnO iii BnO O iv BnO OH
BnO BnO O BnO OH

OTIPS OTIPS OTIPS
140b 142 4c
Scheme 29 Reagents and conditions: (i) CH2¼CHMgBr, THF, 78 1C; (ii) a) TIPSCl,
imidazole, DMF; b) column chromatography; (iii) a) O3, CH2Cl2, 78 1C; b) Me2S; iv) SmI2,
78 1C; then sat. aq. NaHCO3, 78 1C to 20 1C.
designed for the preparation of myo- and chiro-inositols from D-xylose.69

Dienols 140a and 141, prepared from D-xylose derivative 139,70 were sily-
lated and the silylated mixture was separated by silica gel chromatography
to produce pure 140b. Treatment of 136b with ozone produced the corre-
sponding dialdehyde 142, which upon treatment with 6 equivalents of SmI2
at 78 1C for 20 minutes, followed by dropwise addition of saturated
aqueous NaHCO3 at 78 1C and then warming to 20 1C, resulted in the
formation of diol 4c (Scheme 29).
Carbohydr. Chem., 2012, 38, 263–302 | 281

BnO BnO OH BnO OH
OH
i BnO OH BnO OH
BnO OH +

143 144a (1:1) 145a
ii
HO S S
BnO O BnO O
BnO iii O + BnO O
BnO

146 144b 145b
iv
OH HO HO
OH OH
O v
BnO BnO N3 vi HO NH2
BnO OBn BnO OBn HO OH

147 148 149
Scheme 30 Reagents and conditions: (i) a) (COCl)2, DMSO, THF, NEt3, 60 1C; b) SmI2,
t-BuOH, THF, r.t., 90%. (ii) 1,1-thiocarbonyldiimidazole, toluene, 110 1, 97%. (iii) a) Ac2O,
TMSOTf. b) (EtO)3P, heat. c) NaOMe, MeOH, 82%. (iv) L-DIPT, Ti(Oi-Pr)3, t-BuOOH,
CH2Cl2, 94%. (v) LiN3, NH4Cl, DMF, 125 1C, 89%. (vi) H2, Pd(OH)2, EtOH/THF/TFA,
67%.
One of the strategies used for the preparation of polyhydroxylated ami-

nocyclopentanes from carbohydrates is based on the intramolecular pinacol
coupling of sugar aldehydes and/or ketones and subsequent introduction of
the amino group. Thus, Swern oxidation of diol 143, derived from D-glu-
cose, followed by dropwise addition of the crude mixture into a solution of
samarium diiodide in THF containing t-BuOH, gave a 1:1 mixture of
cyclitols 144a and 145a. From this mixture, a sequence involving formation
of the cyclopentene from the cyclic thiocarbonate, epoxidation, ring open-
ing with an azide and hydrogenolysis afforded the desired amino cyclo-
pentanol 149 (Scheme 30).68a
Another useful procedure for the preparation of polyhydroxylated ami-
nocyclopentanes would be the intramolecular pinacol coupling of glucosa-
mine-derived aldehydes. The oxidation-pinacol coupling sequence was used
for the preparation of the 2:1 mixture of amino cyclopentitols 151 and 152
from glucosamine-derived diol 150. The protocol consisted of a one-pot
Swern oxidation followed by SmI2-mediated pinacol coupling (Scheme 31).71
2.4.2 Intramolecular pinacol coupling of aldehydes or ketones and imino
derivatives. The tandem oxidation-pinacol coupling reaction has been
widely used for the construction of the cycloalkane ring from carbohydrate
diols. As the nitrogenated function is not present in the newly created
bonds, it either has to be present in the sugar skeleton or it has to be
introduced by functionalisation of the cycloalkane ring. However, the
process whereby a carbonyl is coupled to an imino derivative would allow
282 | Carbohydr. Chem., 2012, 38, 263–302

BnO BnO OH BnO OH
OH i BnO OH
D-Glucosamine BnO OH + BnO OH
BnO NHBn BnO NHBn BnO NHBn

150 151 152
Scheme 31 Reagents and conditions: (i) a) (COCl)2, DMSO, THF, NEt3, 60 1C; b) SmI2,
t-BuOH, THF, r.t., 82%.
O O OH NH
2
O OBn i
D-Mannose O N O
OH
BnO OAc BnO
153 154
Scheme 32 Reagents and conditions: (i) a) SmI2, t-BuOH, 30 1C; b) H2O; c) LiOH, 98%
(three steps, one pot).
the creation of the carbocyclic ring with concomitant formation of the 1,2-
aminoalcohol functionality and accordingly, the samarium diiodide medi-
ated intramolecular pinacol coupling of oxime ethers and aldehydes or
ketones would allow the direct formation of 2-aminocycloalkan-1-ols.72
Chiara et al. extensively studied the application of this strategy to the
preparation of aminocyclopentitols from carbohydrates.73 In a recent
example, they synthesised a new trehazolin analogue 154 from carbohydrate
precursors, by a highly efficient route based on their previously developed
ketone/oxime ether reductive carbocyclisation reaction for the construction
of the cyclitol ring.74 Starting from D-mannose, condensation with O-ben-
zylhydroxylamine and subsequent oxidation of the released hydroxyl group,
afforded 153, a suitable substrate for the samarium diiodide-promoted
cyclization to cyclopentane 154 (Scheme 32).
The pinacol coupling is highly diastereoselective, with the trans-1,2-
aminoalcohols formed preferentially, because formation of cis-1,2-ami-
noalcohols is disfavoured by 1,3 diaxial strain in the transition state. As
sometimes the preparation of amino alcohols with cis stereochemistry is
required, some protocols were developed aimed at changing the stereo-
selectivity of the reaction in order that cis-1,2-aminoalcohols could pre-
ferentially be obtained.75 A recent strategy developed to overcome the 1,3
diaxial strain and hence obtain cis-amino alcohols is the intramolecular
tethering of the C¼N group to the vicinal hydroxyl group.76 Thus, treat-
ment of 1,2-cyclic carbonate 155a with hydrazine produced hydrazone 156,
which upon oxidation gave substrate 157. Treatment of 157 with samarium
diiodide gave the expected cyclopentitol in 65% yield as a 7:1 mixture of
isomers 158 and 159 (Scheme 33).
2.5 Ring-closing metathesis

Metathesis reactions have been widely used with carbohydrate-related
substrates for the synthesis of new carbohydrate derivatives, heterocycles,
glycomimetics or glycoconjugates compounds.77 With regard to carbasugar
Carbohydr. Chem., 2012, 38, 263–302 | 283

BnO BnO BnO
O i OH O
ii
BnO O BnO N BnO N
NH NH
BnO O O BnO O BnO O
O O
155a 156 157
iii
BnO OH BnO OH
H H
BnO N BnO N
NH + NH
BnO O BnO O
O O
158 159
Scheme 33 Reagents and conditions: (i) NH2NH2 HCl, DIEA, EtOH, 68%; (ii) Dess-Martin,
CH2Cl2, 51%; (iii) SmI2, t-BuOH, THF, 30 1C, 65%.
OH OR6
HO R4O
HO OH R3O OR1
OH OR2
5a-carba-α
α-D-glucopyranose 80b
RCM
O OH
R 4O OH
R4O
R 3O OR1
R 3O OR1
OR2
OR2
D-glucopyranose 160
Scheme 34 Retrosynthetic analysis of carbasugar from carbohydrate.
synthesis, the most applied methodology so far, is the ring-closing

metathesis (RCM) reaction.78 This is due to the fact that it permits the easy
formation of the carbasugar’s ring from a diene acyclic precursor 160, which
is normally obtained by simple modifications to the equivalent carbohy-
drate derivative (Scheme 34).
Many different diene derivatives have been developed as precursors for
the synthesis of carbasugars by means of RCM reactions several general
reviews on synthesis of carbasugars have been published in the recent years
marking a large emphasis on the applications of this methodology for the
preparation of the cycloalkane ring.79 However, this synthetic strategy is of
great scope and today is still very widely used to carry out the synthesis of
new carbasugars.
The most challenging step in the preparation of carbasugars by RCM
from carbohydrate-based dienes is the introduction of double bonds in
the starting structure. We will see below the most actual approaches for the
preparation of carbohydrates-based dienes and their application in the
synthesis of carbasugar by mean of RCM reactions.
284 | Carbohydr. Chem., 2012, 38, 263–302

O O
i OMe CHO
HO OH I ii
BnO OBn
HO OH BnO OBn
D-xylose 161 162
CO2Et
In O
H Br CO2Et
O 163
Bn H OBn iii
166
HO
CO2Et CO2Et
O
OH OH iv OH
HO OH BnO OBn BnO OBn

(+)-cyclophelitol 165 164
Scheme 35 Reagents and conditions: (i) a) MeOH, HCl(aq), 50 1C, b) I2, PPh3, imidazole, THF,
65 1C, c) BnOC(NH)Cl3, TfOH, dioxane, rt; (ii) Zn, THF, H2O, 40 1C, sonication; iii) 163, In,
H2O, La(OTf)3, 48 h; (iv) Grubbs’ catalyst II, CH2Cl2.
I O OMe
i
HO HO
O O
HO OH HO OH
167a 168 169
ii ii
HO OH HO OH
HO iii HO HO iii HO
HO OH HO OH HO OH HO OH
170 5b 171 172
Scheme 36 Reagents and conditions: (i) Zn or In, allyl bromide, THF, H2O, sonication; (ii)
Grubb’s catalyst, CH2Cl2, 91%; (iii) OsO4, CH2Cl2.
One classical transformation in carbohydrate chemistry to introduce

terminal olefins in the non-reducing end is the zinc-mediated fragmentation
of primary iodo-derivatives originally developed by Bernet and Vasella.80
Madsen et al. used this protocol in 2005 for the synthesis of (þ)-cyclo-
phelitol (Scheme 35).81 Thus, a Zn-promoted reductive elimination of iodo-
derivative 161 provided compound 162, with the terminal carbon–carbon
double bond. Consequently, an indium-mediated allylation at the reducing
end of the carbohydrate, produced diene 164, which after reaction with a 2nd
generation Grubbs’ catalyst, allowed the construction of the cyclohexene
ring of compound 165.
This protocol has proven to be very useful for the synthesis of carbohy-
drate-based dienes, and was used for the direct preparation of similar dienes
from o-iodo glycosides (Scheme 36).82 In this one-pot reaction, the resulting
Carbohydr. Chem., 2012, 38, 263–302 | 285

aldehyde was trapped by allyl bromide in a Barbier-type alkylation and zinc
serves a dual function, promoting both the reductive elimination and acti-
vation of allyl bromide for the allylation. This domino zinc-mediated
transformation would convert the iodo glycosides directly into functiona-
lized dienes that can be subsequently converted into carbocycles by RCM.
For example, a domino reaction with 5-iodo-ribofuranosides as 167a in the
presence of zinc or indium and allyl bromide under sonication gave full
conversion and very high yield of the desired dienes 168 and 169. A RCM
reaction gave cyclohexene 5b and hydroxylation of its carbon-carbon
double bond provided carbasugar 170. On the other hand, diene 165 was
converted into carbasugar 172 via cyclohexene 171.83
Moreover, the intermediate aldehyde obtained in this process can be
intercepted with an amine prior to the allylation. The allylation then takes
place on the formed imine and results in the introduction of an amino group
in the final product. Thus, treatment of 173 þ 167a with zinc, benzylamine
and allyl benzoate bromide succeeded in giving the amino diene 175 in
(Scheme 37).
This procedure was recently used by Madsen et al. to develop the synthesis
of the diene 176a, which was easily transformed by RCM into the seven-
membered ring of carbasugar 177 (Scheme 38), the main intermediate in the
O OMe
I i iii OBz
OBz O
O
O O O
O NAcBn
NRBn
173+167a 174a: R=H 175
ii 174b: R=Ac
Scheme 37 Reagents and conditions: (i) Zn, TMSCl, THF, ultrasound, 40 1C, then BnNH2,
3 Å MS, THF, rt, then Zn, BzOCHCHCH2Br, THF, ultrasound, 40 1C; (ii) Ac2O, Et3N,
DMAP, CH2Cl2, 40 1C; (iii) (C3H4N2Mes2)Cl2Ru¼CHC6H4OC3H7, toluene, 80 1C.
O OMe
R
ii iv
BnO NZBn
BnO OPMB BnO NRBn
OBn BnO OPMB
BnO OPMB
177
155b: R=OH 176a: R=H

i iii 176b: R=Z
104b: R=I
HO
NH
HO
HO
calystegine A3
Scheme 38 Reagents and conditions: (i) I2, PPh3, imidazole, THF; (ii) Zn, TMSCl, THF,
sonication, then BnNH2, then CH2¼CHCH2Br, sonication; (iii) ZCl, KHCO3, H2O, CH2Cl2;
(iv) Grubbs’ catalyst II, CH2Cl2.
286 | Carbohydr. Chem., 2012, 38, 263–302

O OH
I
+ Br OBz i O OBn
O O
O OH
167b 178 179
ii
O O
O
HO steps steps
O OBn
HO HO
OH
OH OH
gabosine A 180 gabosine N
Scheme 39 Reagents and conditions: (i) Zn, THF, sonication, then 178, sonication; (ii) Grubbs’
catalyst II, CH2Cl2.
OH
O O OBz OH
HO i BnO ii BnO
HO OH BnO OBn BnO OBn

OH OBn OBn
181 182 183
iii
OH OH
BnO BnO
+ iv BnO OH
BnO OBn BnO OBn
BnO OBn
OBn OBn
OBn
185 186 184
Scheme 40 Reagents and conditions: (i) a) 2-bromoethanol, 89%; b) Pd/C, H2, NaHCO3, H2O,
50 1C; c) BnBr, NaH, DMF, 74% from 181; d) AcOH, HCl(aq), 60 1C, 85%; e) BzCl, py, 86%;
(ii) a) Ph3PCH3Br, t-BuOK, PhMe, 0 1C; b) NaOMe, MeOH, 84%; (iii) a) (COCl)2, Me2SO,
CH2Cl2, 78 1C, then Et3N, 78 1C to rt; b) vinylmagnesium chloride, THF, 0 1C, 76%;
(iv) Grubbs’ catalyst II, toluene, 60 1C; 28% 185 þ 59% 186.
synthesis of calystegine A3.84 A similar strategy was also used for the synthesis
of diene 179, the RCM precursor of gabosine A and gabosine N (Scheme 39).85
Another typical alternative to introduce two carbon-carbon double bonds
into a carbohydrate molecule deals with the employment of the Wittig
reaction and the asymmetric addition of proper Grignard reagents to the
conveniently functionalised carbohydrate molecule. In this way, an inter-
esting synthesis of two carbocyclic mannose mimics 185 þ 186 was devel-
oped from D-fructopyranose 181 through a synthetic pathway where the
olefinic double bonds were sequentially introduced by a Wittig olefination
and an addition of vinylmagnesium chloride (Scheme 40).86
A similar protocol was recently used by Frigell et al. during the synthesis
of compound 189 (Scheme 41), the RCM diene precursor of 4a-carba-a-
galactofuranose 191, from the D-glucose derivative 187.87
Carbohydr. Chem., 2012, 38, 263–302 | 287

BnO BnO BnO
O OH OH
BnO i BnO ii BnO
OH OH
BnO OH BnO OH BnO OH
187 188 189
iii
HO BnO
OH OH
HO iv BnO
HO OH BnO OH
191 190
Scheme 41 Reagents and conditions: (i) vinylmagnesium chloride, THF, rt, 17 h, 87%; (ii) a) 2-
methoxypropene, PPTS, CH2Cl2, rt, 15 min, 94%; b) DMSO, (COCl)2, CH2Cl2, Et3N, rt, 80%;
c) Ph3PMeBr, n-BuLi, toluene, rt, 81%; d) AcOH, H2O, 2 h, quant.; (iii) Grubbs’ catalyst II,
toluene, 48 h; 87%; (iv) H2, Pd/C, EtOAc, EtOH, rt, 1.5 h, quant.
O OH OH OMOM
TBSO i TBSO ii HO
O O O O O O
100b 192 193
iii
HO
O v O
iv
HO HO O HO
OH OH
O
OH OH
(+)-gabosine O 194 (+)-gabosine N
Scheme 42 Reagents and conditions: (i) Ph3P¼CH2, THF, 4 h, 76%; (ii) a) MOMCl, DIPEA,
DMAP, CH2Cl2, 12 h, 93%; b) TBAF, THF, 4 h, 95%; c) (COCl)2, DMSO, CH2Cl2, Et3N, 2 h;
d) 2-bromopropene, CrCl2, NiCl2, DMF, 12 h, 72% or 2-bromopropene, Mg, THF, 4 h, 85%;
(iii) Grubbs’ catalyst II, toluene, 12 h, 85%; (iv) a) PDC, CH2Cl2, 4 Å MS, 12 h, 82%; b)
Amberlyst, THF/H2O, 5 h, 75%; (v) a) PDC, CH2Cl2, 4 Å MS, 12 h, 82%; b) H2, Pd/C, MeOH,
1 h, 95%; c) Amberlyst, THF/H2O, 5 h, 85%.
In this process, both carbon-carbon double bonds were sequentially

introduced in the glucose unit. First, the asymmetric addition of vinylmag-
nesium chloride to hemiacetal 187 provided compound 188 in good yield and
diastereoselectivity. Then, after proper protection, the oxidation of the
hydroxyl group in position C-4 followed by a Wittig olefination produced
diene 189. This compound undergoes RCM with the 2nd generation Grubbs’
catalyst providing cyclopentene 190 in excellent yield, which after hydro-
genation with Pd yielded 4a-carba-a-D-galactofuranose 191.
In a similar way, Rao et al. obtained from D-ribose the carbasugar 194,
a key compound in the synthesis of natural products (þ)-gabosine N and
(þ)-gabosine O (Scheme 42) from D-ribose.88 Unlike the previous case, the
Wittig olefination was employed first in the transformation of the reducing
288 | Carbohydr. Chem., 2012, 38, 263–302

end of the furanoside 100b. The primary hydroxyl group of compound 192
was then oxidised and submitted to an asymmetric addition of an organo-
metallic derivative of 2-bromopropene. The obtained diene 193 was then
efficiently transformed in carbasugar 194 by mean of a 2nd generation
Grubbs’ catalysed RCM reaction.
Another different protocol that allows the preparation of complex RCM
diene precursors from commercially available carbohydrates is based on the
asymmetric aldol reaction of ketones with acrylates. This procedure was
recently used by several authors during the synthesis of natural products
(þ)-pericosine A, (þ)-pericosine B and (þ)-pericosine C as stated in
Scheme 43, which involved a synthetic pathway where the asymmetric
addition of methyl acrylate takes place during the transformation of deri-
vative 196 to diene 197. This compound easily undergoes the formation of a
cyclohexene ring via RCM reaction.89
Finally, another novel alternative for introducing carbon-carbon double
bonds into a carbohydrate-based substrate can be carried.90 In this way,
Betkekar et al. recently described an elegant approach to the 4-methylene-2-
cyclohexenone subunit present in otteliones and loloanolides derivatives, by
way of a strategy that involved a tandem enyne/ring closing metathesis as
the key reaction (Scheme 44).91
Therefore, the addition of lithium acetylene to the carbonyl group of
gluco-D-furanose derivative 67 provided in good yield and diastereoselec-
tivity the compound 199. Then, a terminal double bond was introduced by
the epoxidation of the 5,6-diol system and the regioselective opening of the
epoxide moiety. This was developed by lithium trimethylsilane, producing
enyne 200 which was transformed in cyclohexenol 202 through an efficient
tandem enyne/ring closing metathesis with a 2nd generation Grubbs’ cata-
lyst in the presence of ethylene. The tandem enyne/ring closing metathesis
takes place first by an enyne metathesis reaction of the alquine moiety of 200
MeO2C
HO O
HO
i OMOM ii HO OMOM
O O O O O O
195 196 197
iii
CO2Me
CO2Me CO2Me
HO
MeO MeO
+ iv
O OMOM
HO OH HO OH
O
OH OH
(+)-pericosine B (+)-pericosine C 198
Scheme 43 Reagents and conditions: (i) NaBH4, MeOH, 97%; (ii) a) PivCl, Py, CH2Cl2, 51%;
b) MOMCl, DIPEA, CH2Cl2, 89%, c) NaOMe, MeOH, 95%; d) CrO3, Py, Ac2O, CH2Cl2,
e) methyl acrylate, DABCO, DMF; (iii) Grubbs’ catalyst II, toluene, 86%; (iv) a) Ag2O MeI,
THF, Me2S, 91–95%, b) TFA, MeOH, 81–88%.
Carbohydr. Chem., 2012, 38, 263–302 | 289

O O
O O O
O O i O O ii HO O
O O O O
OH OMe
67 199 200
iii
O OH OH
O O O
O
O iv O
OMe O OMe O O
OMe
203 202 201
Scheme 44 Reagents and conditions: (i) acetylene, n-BuLi, THF, 78%; (ii) a) NaH, MeI, THF
96%, b) AcOH, H2O, 96%, c) p-TsCl, py, d) NaOMe, MeOH, 67%, e) (CH3)3Si, n-BuLi, THF,
95% (iii) Grubbs’ catalyst II, ethylene, CH2Cl2, 83%; (iv) DMP, CH2Cl2 91%.
with one equivalent of ethylene, providing the triene intermediate 201. This
compound thenundergoes an RCM reaction which generates the cyclo-
hexene system of compound 203.
2.6 Cycloaddition reactions

Intramolecular cycloaddition reactions of 1,3-dipoles have been used for the
construction of carbocyclic natural products, including carbohydrate
derivatives.92
Thus, the [3 þ 2] cycloaddition of oximes or nitrones bearing a double
bond was used for the formation of the cyclopentane ring of several bio-
logically relevant 1-amino-2-hydroxymethyl-cyclopentitols.93 For example,
alk-5-enyl aldehyde 204 was easily transformed into the corresponding
oxime, which upon thermolysis resulted in the isolation of hexahydro-1H-
cyclopent[c]isoxazol 205 in good yield, via intramolecular oxime olefin
cycloaddition. Catalytic hydrogenation afforded aziridine 206a, which
finally gave 206b after removal of the protecting groups (Scheme 45).94
Recently, an intramolecular ketonitrone-olefin cycloaddition reaction
was used for the synthesis of aminocyclopentitol 209 (Scheme 46).95
Using a 1,3-dipolar reaction as the key step, Gallos and co-workers
synthesised ent-gabosine E. Condensation of N-methylhydroxylamine with
lactol 210a led to the formation of cycloadducts 211 and 212 in a 2:1 ratio
(Scheme 47), via the corresponding nitrone.96 The major diastereoisomer
211 was then converted to ent-gabosine E in a five further five steps
sequence, in which the isoxazolidine was reductively opened and the amine
eliminated to give the olefin in the desired product.
Nitronates and nitrile oxides can also undergo [3 þ 2] cycloadditions,97
allowing the construction of cycloalkane rings when the reaction occurs
intramolecularly. This strategy was used for the carbocyclisation of sugar
derivatives and for the preparation of pseudo-carbasugars.98 A repre-
sentative example is the synthesis of cyclophellitol 219, reported in 2003.99 A
Henry reaction between D-glucose derived enal 106 and nitromethane
290 | Carbohydr. Chem., 2012, 38, 263–302

BzO OH
OBz OTs
O i ii BzO
BzO O
N N
OBz Bn
TsO Bn BzO
204 205 206a
iii, iv
OH
HO
NH
HO
206b
Scheme 45 Reagents and conditions: (i) BnNHOH.HCl, CaCO3, toluene, 50 1C; (ii) H2, Ni
Raney, EtOAc/EtOH 3:1; (iii) NaOMe, MeOH; (iv) 10% Pd/C, H2, EtOH.
BnO
BnO Bn H2N OH
O OBn
i N ii HO
BnO BnO O OH
OBn OBn HO
BnO OH
207 208 209
Scheme 46 Reagents and conditions: (i) BnNHOH.HCl, CaCO3, toluene, 50 1C; (ii) H2, Ni
Raney, EtOAc/EtOH 3:1; (iii) NaOMe, MeOH; (iv) 10% Pd/C, H2, EtOH.
OH OH
H H
O BnO BnO
i
BnO OH O + O
BnO N BnO N
BnO OBn H Me H Me
OBn OBn
210a 211 212
O OH
HO
HO
OH
ent-gabosine E
Scheme 47 Reagents and conditions: (i) MeNHOH.HCl, NaOEt, EtOH, 20 1C, 80%.
provided nitro olefins 213 and 214, which upon treatment with chloro-
trimethylsilane may lead to cycloadduct 216 by a sequence including a
spontaneous nitronate cycloaddition of the resulting intermediate 215.
In situ treatment of 216 with p-TsOH finally afforded bicyclic compound
217. Silylation, and Mo(CO)6-mediated reductive N-O bond cleavage
afforded the product 218 was easily converted into cyclophelitol 219 in five
steps (Scheme 48).
Carbohydr. Chem., 2012, 38, 263–302 | 291

NO2 NO2
CHO OH OH
i
+
BnO OBn
BnO OBn BnO OBn
OBn
OBn OBn
106 213 214
ii
BnO BnO
O
O
BnO N BnO N
OTMS
OTMS
BnO OTMS BnO OTMS
216 215
iii
BnO BnO CH2OH HO CH2OH

O
iv
BnO N BnO O HO
O
BnO OH BnO OTBS HO
217 218 219
Scheme 48 Reagents and conditions: (i) CH3NO2, TMG, THF, 58% of 107a; (ii) TMSCl,
Et3N, DMAP, THF; (iii) p-TsOH, THF, 55%. (iv) a) TBSOTf, Et3N, CH2Cl2, 95%; b)
Mo(CO)6, MeCN, 90 1C.
HO O O
HON N N
i OH OH
O O O O
O O
220 221 222
ii
HO
iv O OH iii O OH
O OH
O O O O
HO OH
Gabosine O 224 223
Scheme 49 Reagents and conditions: (i) Chloramine T, silica gel, EtOH, 79%, a:b 4.6:1; (ii) H2,
Raney Ni, AcOH, 97%; (iii) Martin’s sulfurane, THF, 78 1C; (iv) a) H2, Raney Ni, EtOH/
H2O/1,4-Dioxane; b) TFA/H2O, 89% (three steps).
Shing and coworkers reported that oxidation of oxime 220 with chlor-
amine T (TolSO2NCl–Na þ ) provided the intermediate nitrile oxide 221 that
spontaneously underwent a cycloaddition to give the bicyclic adduct 222
(Scheme 49).100 The dihydroisoxazole ring was cleaved with Raney nickel
and hydrogen in acetic acid to give the b-hydroxy ketone 223. Regioselective
dehydration of the primary alcohol with Martin’s sulfurane afforded enone
292 | Carbohydr. Chem., 2012, 38, 263–302

224. Catalytic hydrogenation with Raney-Ni followed by hydrolysis of the
acetonide gave natural gabosine O.
2.7 Rearrangement of cyclic sugars

The rearrangement of cyclic sugars allows their direct conversion into
carbocycles. This straightforward approach to carbasugars relies on the
structure of sugar enol-ethers, which inherently carry both masked
nucleophilic and electrophilic functionalities.101
2.7.1 The Ferrier-II rearrangement. The first breakthrough for this class
of reaction was the mercury(II)-mediated rearrangement of hex-5-enopyr-
anoside derivatives, known as the Ferrier (II) rearrangement.102 The
representative example included in Scheme 50 show the mechanism for this
transformation which involves a regioselective hydroxymercuration of hex-
5-enopyranoside 225a to give an unstable acetal 226 that decomposes to the
ketoaldehyde 227 via the loss of methanol.103 A subsequent intramolecular
aldol-cyclisation affords cyclohexanone 228a in high yields, as a single
diastereomer (as in most cases). The stereochemistry of the newly formed
stereogenic center linked to the hydroxyl group is dependent on the ste-
reochemistry of the C-3 substituent: both groups are usually trans.
Since then, this reaction has provided a practical route to a large variety
of bioactive substances such as aminocyclitols,104 pseudosugars,105 inosi-
tols106 and other complex hexitols.107
Despite the usefulness of this procedure, an important drawback is the
toxicity of mercury, which prompted the development of nontoxic alter-
natives. Thus, Adam reported a similar rearrangement of amino-6-deoxy-
hex-5-enopyranoside derivatives mediated by PdCl2 or Pd(OAc)2 in the
presence of aqueous sulfuric acid,108 which was later applied to several 6-
deoxyhex-5-enopyranosides.109 For example, when benzoyl protected
gluco-hexenopyranoside 225b was treated with a catalytic amount of PdCl2,
cyclohexane 228b was obtained in good yield and excellent diastereoselec-
tivity (a/bW99:1) (Scheme 51).110
HgCl
O OMe ClHg
O OMe O O
HO
BnO OTs
BnO OTs BnO OTs
OBn
OBn OBn
225a 226 227
O OH
BnO OTs
OBn
228a
Scheme 50 Reagents and conditions: HgCl2, acetone, reflux.
Carbohydr. Chem., 2012, 38, 263–302 | 293

O OMe O OH
i
BzO OBz BzO OBz
OBz OBz
225b 228b
Scheme 51 Reagents and conditions: (i) PdCl2, dioxane/H2O 2:1, 60 1C, 68%.
O O
O OAc
AcO O
i ii
BnO OBn
BnO OBn BnO OBn
OBn
OBn OBn
155c + 229 230 231
iii
MeO CH2OH MeO CO2Me

iv
HO OH BnO OBn
OH OBn
233 232
Scheme 52 Reagents and conditions: (i) a) Furan, 4 Å MS, TMSOTf, CH3CN, 40 1C to
20 1C, 71%; b) NaOMe, MeOH, 91%; c) I2, PPh3, imidazole, toluene, 70 1C, 89%; d) NaH, DMF,
72%; (ii) TIBAL, toluene, 50 1C, 83%; (iii) a) MeI, NaH, DMF, 96%; b) O3, CH2Cl2, MeOH,
78 1C, then KHCO3, MeI, DMF, 62%; (iv) a) LiAlH4, Et2O, 76%; b) H2, Pd/C, MeOH, 75%.
Another drawback of the Ferrier (II) reaction is that, despite being a

really efficient way to form six-membered carbocycles, it is unsuitable to
form cyclopentitols.
2.7.2 Endocyclic cleavage induced rearrangement. In order to overcome

the limitations of the Ferrier rearrangement, different alternatives were
investigated.
Prof. Pierre Sinay and co-workers reported an alternative which consists
of a direct conversion of hex-5-enopyranosides into highly functionalised
cyclohexane derivatives by reductive rearrangement induced by triisobuty-
laluminium (TIBAL).111 Later on, Sinay and Sollogoub also demonstrated
that the titanium-based Lewis acid Ti(OiPr)Cl3 was also able to mediate the
rearrangement of O-glycosides under milder conditions.112 For example C-
furyl glycoside, derived from D-glucose (155c þ 229), was converted to a
partially protected carba-a-D-idopyranoside 230 (Scheme 52).113 The furyl
group was used as a masked form of the hydroxymethyl moiety at C-5.
A remarkable feature of this type of rearrangements is the retention of the
anomeric configuration. Accordingly, this methodology was employed in
the synthesis of carbadisaccharides from disaccharides.114
On the other hand, modified carbohydrates, where the terminal carbon
has been elaborated into a vinyl unit, are useful precursors of cyclopentanes
via an intramolecular reaction between aldehyde and an allylmetal species
generated in situ.115 Thus, a zirconium-mediated ring-contraction is induced
294 | Carbohydr. Chem., 2012, 38, 263–302

OHC O OMe O OMe
i
PMBO OBn PMBO OBn
OPMB OPMB
116b 210b
ii
PMBO OH + PMBO OH
PMBO OBn PMBO OBn

234a 235a
Scheme 53 Reagents and conditions: (i) CH3PPh3Br, t-BuOK, THF; (ii) ‘‘Cp2Zr’’/THF then
BF3 Et2O, 78 1C to rt.
OH OH
O OAc
i
+
BnO OBn BnO OBn
BnO OBn
OBn OBn OBn
210c 234b 235b
Scheme 54 Reagents and conditions: (i) SmI2, Pd(Ph3)4, THF, 80 1C, 82% 234b and 12% 235b.
by sequential treatment of vinyl carbohydrate derivatives with ‘‘Cp2Zr’’ and

BF3 OEt2.116 A recent example is the preparation of vinyl cyclopentanes
234a and 235a from sugar aldehyde 116b, in which the key step is a ring
contraction carried out by treating 210b with Cp2Zr-(n-Bu)2 (prepared
in situ from zirconocene dichloride and 2 equiv of n-butyllithium), followed
by boron trifluoride etherate in THF (Scheme 53).117
Samarium(II) iodide in the presence of Pd(0) can also promote this
transformation. An example is the transformation of 5-alkenylpyranoside
210c into homoallyl cyclopentanols 234b and 235b (Scheme 54).118 The first
step is the formation of an allenylpalladium complex, followed by reduction
by SmI2 and subsequent intramolecular allylation of the generated carbo-
nyl-tethered allenylsamarium.
Comparing both processes, they differ in the stereochemical outcomes
that are possible in the final cyclopentanes depending on the particular
metal employed. Thus, allylzirconium intermediates lead to a syn relation-
ship between vinyl and hydroxyl groups in the cyclised products, whereas
allylsamarium intermediates preferentially yield the corresponding anti
products.119
Recently, the possibility of using the Pd(0)/Et2Zn system in this transfor-
mation was also investigated.120 Upon treatment of vinyl carbohydrate 210c
with Et2Zn/Pd(PPh3)4/ZnCl2 the cyclopentane products 236, 237, 238 and
239 were obtained in good yields and diastereoisomeric ratio (Scheme 55).
2.7.3 Claisen rearrangement. Recently, a synthesis of antiviral Tamiflu
from cheap and abundantly available D-glucal (240a) was described,
Carbohydr. Chem., 2012, 38, 263–302 | 295

OH OH
+
BnO OBn BnO OBn
OBn OBn
O OAc 236 237
i
BnO OBn
OH OH
OBn
210c +
BnO OBn BnO OBn
OBn OBn
238 239
Scheme 55 Reagents and conditions: (i) Et2Zn, Pd(0), ZnCl2, THF, r.t., 6% of 224, 6% of 225,
75% of 226 and 13% of 227.
O O O O
HO i HO
ii
HO PMP O PMBO
OH OTBS OTBS
240a 240b 240c
iii
CHO O
iv
OTBS PMBO
OPMB OTBS
242 241
Scheme 56 Reagents and conditions: (i) a) p-anisaldehyde diethyl acetal, PPTS, DMF, 2 h; b)
TBSCl, imidazole, DMAP, DMF, 3 h; (ii) DIBAL-H, dichloromethane, 15 to 0 1C, 2 h, 65%
(3 steps); (iii) Dess-Martin, dichloromethane, 2 h followed by methyltriphenylphosphonium
bromide, nBuLi, THF, 78 to 25 1C, 1 h, 67% (2 steps); (iv) diphenyl ether, 210 1C, 2 h, 88%.
featuring a Claisen rearrangement to obtain the cyclohexene backbone

(Scheme 56). Installation of a 4,6-benzylidene acetal and silylation of the 3-
hydroxyl group gave fully protected D-glucal 240b, which underwent
selective opening of the benzylidene acetal with DIBAL-H in dichlor-
omethane to give the free primary alcohol 240c. The primary hydroxyl
group in 240c was oxidised to the aldehyde by using Dess–Martin period-
inate and then subjected to Wittig methylenation to give terminal olefin 241
in 67% yield. The next step was the critical Claisen rearrangement reaction,
which allowed ready access to carbocycle 242.121
Acknowledgements
We thank the Spanish Ministry of Science and Innovation (CTQ2009-
08490) and by the Xunta de Galicia (Research Project CN2011/037) for
financial support and the University of Aveiro, the Fundação para a Ciência
e a Tecnologia (FCT) and the Fundo Europeu de Desenvolvimento
Regional (FEDER) for funding the Organic Chemistry Research Unit. R.S.
296 | Carbohydr. Chem., 2012, 38, 263–302

thanks the FCT for a ‘‘Investigador Auxiliar’’ position. A.M.E. thanks the
Spanish Ministry of Science and Innovation for an FPU grant.
References
1 (a) O. A. Arjona, M. Gómez, J. C. Lopez, J. Plumet, Chem. Rev., 2007, 107,
1919–2036.
2 (a) E. B. d. Melo, A. d. S. Gomes, I. Carvalho, Tetrahedron, 2006, 62, 10277–
10302; (b) P. Vogel, Chimia, 2001, 55, 359–365.
3 G. E. McCasland, S. Furuta, L. J. Durham, J. Org. Chem., 1966, 31, 1516–
1521.
4 (a) G. E. McCasland, S. Furuta, L. J. Durham, J. Org. Chem., 1968, 33 (7),
2835–2841; (b) G. E. McCasland, S. Furuta, L. J. Durham, J. Org. Chem.,
1968, 33, 2841–2844.
5 T. Suami, S. Ogawa, Adv. Carbohydr. Chem. Biochem., 1990, 48, 21–90.
6 (a) P. Sears, C. H. Wong, Angew. Chem. Int. Ed., 1999, 38, 2300–2324; (b) J.
Sardinha, A. P. Rauter, M. Sollogoub, Tetrahedron Lett., 2008, 49, 5548–
5550.
7 T. W. Miller, B. H. Arison and G. Albers-Schonberg, Biotechnol. Bioeng.,
1973, 15, 1075–1080.
8 (a) T. Posternak, Les cyclitols, Hermann, Paris, 1962; (b) D. J. Cosgrove,
G. C. J. Irving, Inositol Phosphates, Elsevier, Amsterdam, 1980; (c) B. Kilbas,
M. Balci, Tetrahedron, 2011, 67, 2355–2389.
9 (a) M. Balci, Y. Sutbeyaz, H. Secen, Tetrahedron, 1990, 46, 3715–3742;
(b) M. Honzumi, K. Hiroya, T. Taniguchi, K. Ogasawara, Chem. Commun.,
1999, 1985–1986.
10 (a) K. Tatsuta, ‘Carbohydrate Mimics’, Wiley-VCH Verlag GmbH, Wein-
heim, 1998; (b) K. Tatsuta, Pure & Appl. Chem., 1996, 68, 1341–1346.
11 (a) Y. Kameda, N. Asano, M. Teranishi, M. Yoshikawa, K. J. Matsui,
J. Antibiot., 1981, 34, 1237–1240; (b) N. Asano, M. Takeuchi, K. Ninomiya,
Y. Kameda, K. Matsui, J. Antibiot., 1984, 37, 859–867; (c) S. Ogawa, Y.
Miyamoto, A. Nakajima, Chem. Lett., 1989, 725–726; (d) S. Ogawa, A.
Nakajima, Y. Miyamoto, J. Chem. Soc., Perkin Trans., 1991, 1, 3287–3290.
12 S. Horii, T. Iwasa and Y. Kameda, J. Antibiot., 1971, 24, 57–58.
13 (a) X. Chen, Y. Fan, Y. Zheng, Y. Shen, Chem. Rev., 2003, 103, 1955–1977;
(b) T. Mahmud, Nat. Prod. Rep., 2003, 20, 137–166.
14 (a) H. O. L. Fischer, G. Dangschat, Helv. Chim. Acta, 1934, 17, 1196–1200;
(b) H. O. L. Fischer, G. Dangschat, Helv. Chim. Acta, 1934, 17, 1200–1204;
(c) H. O. L. Fischer, G. Dangschat, Helv. Chim. Acta, 1935, 18, 1204–1206;
(d) H. O. L. Fischer, G. Dangschat, Helv. Chim. Acta, 1935, 18, 1206–1213;
(e) H. O. L. Fischer, G. Dangschat, Helv. Chim. Acta, 1937, 20, 705–716.
15 A. Barco, S. Benetti, C. D. Risi, P. Marchetti, G. P. Pollini and V. Zanirato,
Tetrahedron, Asymm., 1997, 8, 3515–3545.
16 (a) J. Marco-Contelles, Eur. J. Org. Chem., 2001, 1607–1618. (b)
Y. Kobayashi, ‘Glycoscience, Chemistry and Chemical Biology’, Springer-
Verlag, Berlin, 2001. (c) S. Atsumi, K. Umezawa, H. Iinuma, H. Naganawa,
H. Nakamura, Y. Iitaka, T. Takeuchi, J. Antibiot., 1990, 43, 49–53.
17 S. Atsumi, H. Iinuma, C. Nosaka and K. Umezawa, J. Antibiot., 1990, 43,
1579–1587.
18 Y. Nobuji, C. Noriko, M. Takashi, U. Shigeru, A. Kenzon, I. Michaki, Jpn.
Kokai Tokkyo Koho, Vol. 06306000, JP, 1994.
19 A. M. Estévez, R. J. Estévez, J. C. Estévez, Mini-rev. Med. Chem. 2012,
in press.
Carbohydr. Chem., 2012, 38, 263–302 | 297

20 D. Bradley, Nat. Rev. Drug Discov., 2005, 4, 945–946.
21 T. Iwasa, H. Yamamoto and Y. Kameda, J. Antibiot., 1970, 32, 595–602.
22 A. Berecibar, C. Grandjean and A. Siriwardena, Chem. Rev., 1999, 99,
779–844.
23 H. Tsunoda, S. Sasaki, T. Furuya and S. Ogawa, Liebigs Ann. Chem., 1996,
159–165.
24 (a) S. I. Hakomori, Adv. Cancer Res., 1989, 52, 257–331; (b) M. Fukuda,
O. Hindsgaul, IRL Press, Oxford, New York, Tokyo, 1994.
25 R. J. Ferrier, Chem. Rev., 1993, 93, 2779–2831.
26 (a) T. Suami, K. Tadano, Y. Kameda, Y. Iimura, Chem. Lett. 1984, 1919–
1922. (b) K. Tadano, Y. Kameda, Y. Iimura, T. Suami, J. Carbohydr. Chem.
1987, 6, 231–244.
27 K. Tadano, H. Maeda, M. Hoshino, Y. Iimura and T. Suami, Chem. Lett.,
1986, 1081–1084.
28 R. G. Soengas, J. C. Estévez, A. M. Estévez, F. Fernández and R. J. Estévez,
Recent advances in nitrosugar chemistry, Carbohydrate Chemistry, 2009,
173–199.
29 L. Henry, Bull. Soc. Chim. France, 1895, 13, 999–1001.
30 (a) H. B. Hass, M. L. Bender, J. Am. Chem. Soc., 1949, 71, 1767–1769. (b) S. J.
Etheredge, Tetrahedron Lett., 1965, 50, 4527–4530.
31 P. W. Wade, R. M. Giuliano, Nitro Compounds Recent Advances in Synthesis
and Chemistry, 1990, ed. by H. Feuer and A. T. Nielsen, VCH, New York.
32 H. H. Baer, L. Urbas, The Chemistry of the Nitro and Nitroso Groups, Willey
Interscience, New York, 1970, part 2, 75.
33 M. Yoshikawa, B. Cha, T. Nakae and I. Kitagawa, Chem. Pharm. Bull., 1984,
36, 3714–3717.
34 (a) J. Kovar, H. H. Baer, Can. J. Chem., 1973, 51, 1801–1811; (b) J. Kovar,
H. H. Baer, Can. J. Chem., 1973, 51, 2836–2842; (c) M. Yoshikawa, B. Cha,
T. Nakae, I. Kitagawa, Chem. Pharm. Bull., 1984, 36, 3714–3717. (d)
K. Brewster, J. M. Harrison, T. D. Inch, N. Williams, J. Chem. Soc. Perkin
Trans. I, 1987, 21–26; (e) S. D. Gero, D. Horton, A. M. Sepulchre, D. Wander,
Tetrahedron, 1973, 29, 2963–2972; (f) M. Yosikawa, Y. Yokokawa, Y. Inoue,
S. Yamagichi, N. Murakami, Tetrahedron, 1994, 50, 9961–9974; (g) E. J. Corey,
I. Vlattas, K. Harding, J. Am. Chem. Soc., 1969, 91, 535–536; (h) S. Tori,
T. Inokuchi, R. Oi, K. Kondo, T. Kobayashi, J. Org. Chem., 1986, 51, 254–256.
35 R. G. Soengas, J. C. Estévez and R. J. Estévez, Org. Lett., 2003, 5, 4457–4459.
36 L. El Blidi, M. Ahbala, J. Bolte and M. Lemaire, Tetrahedron: Asymm., 2006,
17, 2684–2688.
37 C. Chakraborty, V. P. Vyavahare, V. G. Puranic and D. D. Dhavale, Tetra-
hedron, 2008, 64, 9574–9580.
38 J. M. Otero, F. Fernandez, J. C. Estévez and R. J. Estévez, Tetrahedron:
Asymm., 2005, 16, 4045–4049.
39 R. G. Soengas, J. C. Estévez and R. J. Estévez, Org. Lett., 2003, 5, 1423–1425.
40 (a) R. G. Soengas, M. B. Pampin, J. C. Estévez, R. J. Estévez, Tetrahedron
Asymm., 2005, 16, 205–211. (b) A. M. Estévez, R. G. Soengas, R. Tato, P.
Thomas, J. C. Estévez, R. J. Estévez, Tetrahedron Asymm., 2010, 20, 116–120.
41 (a) H. Paulsen, W. v. Dein, Liebigs Ann. Chem., 1987, 125–131; (b) H. Fukase,
S. Horii, J. Org. Chem., 1992, 57, 3651–3658; (c) S. Cai, M. Stroud, S.
Hakomori, T. Toyokuni, J. Org. Chem., 1992, 57, 6693–6696.
42 (a) T. K. M. Shing, Y. Chen, W.-L. Ng, Tetrahedron, 2011, 67, 6001–6005. (b)
T. K. M. Shing, Y. Chen, J. Org. Chem., 2007, 72, 6610–6613.
43 (a) T. Suami, K. Tadano, Y. Ueno, C. Fukabori, Chem. Lett., 1985, 1557–
1560; (b) K. Tadano, C. Ueno, C. Fukabori, Y. Hotta, T. Suami, Bull. Chem.
298 | Carbohydr. Chem., 2012, 38, 263–302

Soc. Jpn., 1987, 60, 1727–1739; (c) K. Tadano, C. Fukabori, M. Miyazaki, H.
Kimura, T. Suami, Bull.Chem. Soc. Jpn., 1987, 60, 2189–2196.
44 (a) G. Rassu, L. Auzzas, L. Pinna, L. Battistini, F. Zanardi, L. Marzocchi,
D. Acquotti, G. Casiraghi, J. Org. Chem., 2000, 65, 6307–6318; (b) F. Zanardi,
L. Battistini, L. Marzocchi, D. Acquotti, G. Rassu, L. Pinna, L. Auzzas,
V. Zambrano, G. Casiraghi, Eur. J. Org. Chem., 2002, 1956–1965.
45 (a) T. K. M. Shing, H. M. Cheng, Org. Biomol. Chem., 2009, 7, 5098–5102;
(b) T. K. M. Shing, H. M. Cheng, Synlett, 2010, 142–144.
46 (a) T. K. M. Shing, H. M. Cheng, W. F. Wong, C. S. K. Kwong, J. Li, C. B. S.
Lau, P. S. Leung and C. H. K. Cheng, Org. Lett., 2008, 10, 3145–3148.
(b) T. K. M. Shing and H. M. Cheng, Org. Lett., 2008, 10, 4137–4139.
47 (a) K. Tatsuta, M. Takahashi. N. Tanaka, J. Antibiot., 2000, 53, 88–91.
(b) K. Tatsuta, H. Mukai, M. Takahashi, J. Antibiot., 2000, 53, 430–435.
48 D. H. Mac, R. Samineni, J. Petrignet, P. Srihari, S. Chandrasekhar, J. S.
Yadav and R. Grée, Chem. Commun., 2009, 4717–4719.
49 (a) E. J. Corey, D. Seebach, Angew. Chem., Int. Ed. Engl., 1965, 4, 1075–1077;
(b) D. Seebach, N. R. Jones, E. J. Corey, J. Org. Chem., 1968, 33, 300–305.
50 Y.-L. Chen, H. Redlich, K. Bergander and R. Fröhlich, Org. Biomol. Chem.,
2007, 5, 3330–3339.
51 X. Cheng, N. Khan, G. Kumaran and D. R. Mootoo, Org. Lett., 2001, 3,
1323–1325.
52 For the first references of radical cyclization of carbohydrate derivatives
leading to carbasugars see: (a) W. Schmid, G. M. Whitesides, J. Am. Chem.
Soc., 1990, 112, 9670–9671; (b) F. O. Andersson, B. Classon, B. Samuelsson,
J. Org. Chem., 1990, 55, 4699–4704.
53 H. Redlich, W. Sudau, A. Szardenings and R. Vollerthun, Carbohydr. Res.,
1992, 226, 57–78.
54 (a) B. Bernet, A. Vasella, Helv. Chim. Acta, 1979, 62, 1990–2016. (b) B. Bernet,
A. Vasella, Helv. Chim. Acta, 1979, 62, 2400–2410. (c) B. Bernet, A. Vasella,
Helv. Chim. Acta, 1979, 62, 2411–2431. (d) M. Nakane, C. R. Hutchinson, H.
Gollman, Tetrahedron Lett., 1980, 21, 1213–1216. (e) A. Fürstner, D. Jumbam,
J. Teslic, H. Weidmann, J. Org. Chem., 1991, 56, 2213–2217. (f) R. J. Ferrier,
P. J. Prasit, Chem. Soc., Chem. Comm., 1981, 983–985. (g) A. Fürstner,
J. Baumgartner, D. N. J. Jumbam, Chem. Soc., Perkin Trans., 1993, 1, 131–138.
55 J. J. C. Grové, C. W. Holzapfel and D. B. G. Williams, Tetrahedron Lett.,
1996, 37, 1305–1308.
56 J. J. C. Grové, C. W. Holzapfel and D. B. G. Williams, Tetrahedron Lett.,
1996, 37, 5817–5820.
57 J. L. Chiara, S. Martı́nez and M. Bernabe, J. Org. Chem., 1996, 61, 6488–6489.
58 A. Chenede, P. Pothier, M. Sollogoub, A. J. Fairbanks and P. Sinay, J. Chem.
Soc., Chem. Comm., 1995, 1373–1374.
59 J. J. C. Grové and C. W. Holzapfel, Tetrahedron Lett., 1997, 38, 7429–7432.
60 J. Marco-Contelles, P. Gallego, M. Rodrı́guez-Fernández, N. Khiar,
C. Destabel, M. Bernabé, A. Martı́nez-Grau and J. L. Chiara, J. Org. Chem.,
1997, 62, 7397–7412.
61 (a) A. Gómez, E. Moreno, S. Valverde, J. López, Synlett, 2002, 891–894;
(b) A. Gómez, E. Moreno, S. Valverde, J. López, Eur. J. Org. Chem., 2004,
1830–1840; (c) A. Gómez, E. Moreno, S. Valverde, J. López, Tetrahedron
Lett., 2002, 43, 5559–5562.
62 G. M. Robertson, Comprehensive Organic Synthesis, Oxford: Pergamon Press,
1991; vol. 3, 563.
63 (a) J. L. Namy, J. Souppe, H. B. Kagan, Tetrahedron Lett., 1983, 24, 765–
766; (b) J. L. Chiara, W. Cabri, S. Hanessian, Tetrahedron Lett., 1991, 32,
Carbohydr. Chem., 2012, 38, 263–302 | 299

1125–1128; (c) F. Machrouhi, J.-L. Namy, Tetrahedron Lett., 1999, 40, 1315–
1318; (d) R. Annunziata, M. Benaglia, M. Cinquini, F. Cozzi, L. Raimondi,
Tetrahedron Lett. 1998, 39, 3333–3336; (e) R. D. Rieke, S.-H. Kim, J. Org.
Chem. 1998, 63, 5235–5239; (f) S. Talukdar, A. Banerji, J. Org. Chem., 1998,
63, 3468–3470; (g) A. Clerici, L. Clerici, O. Porta, Tetrahedron Lett. 1995, 36,
5955–5958; (h) M. P. Dutta, B. Baruha, A. Baruah, D. Prajapati, J. S. Sandhu,
Synlett, 1998, 857–858.
64 (a) J. L. Chiara, W. Cabri, S. Hanessian, Tetrahedron Lett., 1991, 32, 1125–
1128; (b) J. Uenishi, S. Masuda, S. Wakabayashi, Tetrahedron Lett., 1991, 32,
5097–5100; (c) J. L. Chiara, M. Martı́n-Lomas, Tetrahedron Lett., 1994, 35,
2969–2972; (d) J. P. Guidot, T. Le Gall, Mioskowski, Tetrahedron Lett., 1994,
36, 6671–6672; (e) J. L. Chiara, N. Valle, Tetrahedron: Asymm., 1995, 6, 1895–
1898; (f) J. L. Chiara, N. Valle, Tetrahedron Lett., 1994, 35, 2969–2972. (g) D.
J. Jenkins, B. V. L. Potter, J. Chem. Soc. Perkin Trans., 1998, 1, 41–50; (h)
A. Kornienko, D. I. Turner, C. H. Jaworek, M. d’Alarcao, Tetrahedron:
Asymm. 1998, 9, 2783. (i) T. Sawada, R. Shirai, S. Iwasaki, Tetrahedron Lett.,
1996, 37, 885–886; (j) M. Adinolfi, G. Barone, A. Iadonisi, L. Mangoni,
R. Manna, Tetrahedron, 1997, 11767–11780. (k) M. Carpintero, C. Jaramillo,
A. Fernández-Mayoralas, Eur. J. Org. Chem., 2000, 1285–1296. (l) I. Storch de
Gracia, H. Dietrich, S. Bobo, J. L. Chiara, J. Org. Chem. 1998, 63, 5883–5889.
65 J. L. Chiara, W. Cabri and S. Hanessian, Tetrahedron Lett., 1991, 32,
1125–1128.
66 (a) J. L. Chiara, M. Martı́n-Lomas, Tetrahedron Lett., 1994, 35, 2969–2972.
(b) A. Kornienko, D. I. Turner, C. H. Jaworek, M. d’Alarcao, Tetrahedron:
Asymm., 1998, 9, 2783–2786. (c) D. J. Jenkins, B. V. L. Potter, J. Chem. Soc.
Perkin Trans. 1, 1998, 41–49. (d) J. P. Guidot, T. Le Gall, C. Mioskowski,
Tetrahedron Lett., 1994, 36, 6671–6672.
67 (a) M. Adinolfi, G. Barone, A. Iadonisi, L. Mangoni, R. Manna, Tetrahedron,
1997, 11767–11780. (b) M. Carpintero, C. Jaramillo, A. Fernández-
Mayoralas, Eur. J. Org. Chem., 2000, 1285–1296.
68 (a) I. Storch de Gracia, H. Dietrich, S. Bobo, J. L. Chiara, J. Org. Chem., 1998,
63, 5883–5889. (b) J. L. Chiara, I. Storch de Gracia, A. Bastida, Chem.
Commun., 2003, 1874–1875.
69 G. Luchetti, K. Ding, A. Kornienko and M. d’Alarcao, Synthesis, 2008,
3148–3154.
70 G. Luchetti, K. Ding, M. d’Alarcao and A. Kornienko, Synthesis, 2008,
3142–3147.
71 J. L. Chiara, I. Storch de Gracia and A. Bastida, Chem. Commun., 2003,
1874–1875.
72 (a) H. Miyabe, M. Torieda, T. Kiguchi, T. Naito, Synlett, 1997, 580–582;
(b) H. Miyabe, M. Torieda, K. Inoue, K. Tajiri, T. Kiguchi, T. Naito, J. Org.
Chem., 1998, 63, 4397–4407.
73 (a) J. L. Chiara, J. Marco-Contelles, N. Khiar, P. Gallego, C. Destabel,
M. Bernabé, J. Org. Chem., 1995, 60, 6010–6011; (b) S. Bobo, I. Storch de
Gracia, J. L. Chiara, Synlett, 1999, 1551–1554; (c) I. Storch de Gracia,
S. Bobo, M. D. Martı́n-Ortega, J. L. Chiara, Org. Lett., 1999, 11, 1705–1708.
74 J. L. Chiara, I. Storch de Gracia, A. Garcı́a, A. Bastida, S. Bobo and M. D.
Martı́n-Ortega, ChemBioChem, 2005, 6, 186–191.
75 A. Boiron, P. Zillig, D. Faber and B. Giese, J. Org. Chem., 1998, 63, 5877–
5882.
76 J. L. Chiara and A. Garcı́a, Synlett, 2005, 2607–2610.
77 (a) J. C. López, J. Plumet, Eur. J. Org. Chem., 2011, 1803–1825; (b) J. Li,
D. Lee, Eur. J. Org. Chem., 2011, 4269–4287.
300 | Carbohydr. Chem., 2012, 38, 263–302

78 (a) S. T. Nguyen, L. K. Johnson, R. H. Grubbs, J. Am. Chem. Soc., 1992, 114,
3974–3975; (b) G. C. Fu, S. T. Nguyen, R. H. Grubbs, J. W. Ziller, J. Am.
Chem. Soc., 1993, 115, 9856–9857; (c) P. Schwab, R. H. Grubbs, J. W. Ziller,
J. Am. Chem. Soc., 1996, 118, 100–110.
79 (a) R. Madsen, Eur. J. Org. Chem., 2007, 399–415; (b) J. Plumet, A. M.
Gomez, J. C. López, Mini-Reviews in Org. Chem., 2007, 4, 201–216; (c) V. B.
Kurteva, C. A. M. Afonso, Chem. Rev., 2009, 109, 6809–6857.
80 (a) B. Bernet, A. Vasella, Helv. Chim. Acta 1979, 62, 2400–2411; (b)
M. Nakane, C. R. Hutchinson, H. Gollman, Tetrahedron Lett. 1980, 21, 1213–
1216; (c) A. Fürstner, D. Jumbam, J. Teslic, H. Weidmann, J. Org. Chem.,
1991, 56, 2213–2217; (d) R. J. Ferrier, P. J. Prasit, Chem. Soc., Chem. Comm.,
1981, 983–984; (e) A. Fürstner, J. Baumgartner, D. N. Jumbam, J. Chem. Soc.,
Perkin Trans., 1993, 1, 131–138.
81 F. G. Hansen, E. Bundgaard and R. Madsen, J. Org. Chem., 2005, 70, 10139–
10142.
82 M. Jorgensen, P. Hadwiger, R. Madsen, A. E. Stutz and T. M. Wrodnigg,
Curr. Org. Chem., 2000, 4, 565–588.
83 L. Hyldtoft and R. Madsen, J. Am. Chem. Soc., 2000, 122, 8444–8452.
84 (a) P. R. Skaanderup, R. Madsen, J. Org. Chem., 2003, 68, 2115–2122; (b)
R. N. Monrad, C. B. Pipper, R. Madsen, Eur. J. Org. Chem., 2009, 3387–3395.
85 R. N. Monrad, M. Fanefjord, F. G. Hansen, N. M. E. Jensen and R. Madsen,
Eur. J. Org. Chem., 2009, 396–402.
86 C. Ramstadius, M. Boklund and I. Cumpstey, Tetrahedron: Asymmetry, 2011,
22, 399–405.
87 J. Frigell, L. Eriksson and I. Cumpstey, Carbohydr. Res., 2011, 346, 1277–
1290.
88 J. P. Rao and B. V. Rao, Tetrahedron: Asymmetry, 2010, 21, 930–935.
89 (a) W. H. Kim, J. A. Kang, H. R. Lee, A. Y. Park, P. Chun, B. Lee, J. Kim,
J. A. Kim, L. S. Jeong, H. R. Moon, Carbohydr. Res. 2009, 344, 2317–2321; (b)
S. Tripathi, A. C. Shaikh, C. Chen, Org. Biomol. Chem., 2011, 9, 7306–7308;
(c) Y. S. Reddy, P. Kadigachalam, R. K. Basak, A. P. J. Pal, Y. D. Vankar,
Tetrahedron Lett., 2012, 53, 132–136; (d) C. MuniRaju, J. P. Rao, B. V. Rao,
Tetrahedron: Asymmetry, 2012, doi:10.1016/j.tetasy.2012.01.002.
90 S. T. Diver and A. J. Giessert, Chem. Rev., 2004, 104, 1317–1382.
91 V. V. Betkekar, S. Panda and K. P. Kaliappan, Org. Lett., 2012, 14, 198–201.
92 J. M. B. Adam and I. Coldham, Curr. Org. Synth., 2010, 7, 312–331.
93 (a) H. M. I. Osborn, N. Gemmell, L. M. Harwood, J. Chem. Soc., Perkin
Trans., 2002, 1, 2419–2438; (b) A. E. Koumbis, J. K. Gallos, Curr. Org. Chem.,
2003, 7, 585–628; (c) J. K. Gallos, A. E. Koumbis, V. P. Xinpbaki, C. C.
Dellios, E. Coutoul-Argyropoulou, Tetrahedron, 1999, 55, 15167–15180; (d)
P. J. Dransfield, S. Moutel, M. Shipman, V. Sik, J. Chem. Soc., Perkin Trans.,
1999, 1, 3349–3355.
94 O. L. Lopez, J. G. Fernández-Bolaños, V. H. Lillelund and M. Bols, Org.
Biomol. Chem., 2003, 1, 478–482.
95 Y. S. Reddy, P. Kadigachalam, V. R. Doddi and Y. D. Vankar, Tetrahedron
Lett., 2009, 50, 5827–5830.
96 C. I. Stathakis, M. N. Athanatou and J. K. Gallos, Tetrahedron Lett., 2009,
50, 6916–6918.
97 N. Ono, The Nitro Group in Organic Synthesis; Wiley-VCH: New York, 2001;
pp. 267–274.
98 A. E. Koumbis and J. K. Gallos, Curr. Org. Chem., 2003, 7, 398–426.
99 T. Ishikawa, Y. Shimizu, T. Kudoh and S. Saito, Org. Lett., 2003, 5, 3879–
3882.
Carbohydr. Chem., 2012, 38, 263–302 | 301

100 (a) T. K. M. Shing, W. F. Wong, H. M. Cheng, W. S. Kwok, K. H. So, Org.
Lett., 2007, 9, 753–756. (b) T. K. M. Shing, K. H. So, W. S. Kwok, Org. Lett.,
2009, 11, 5070–5073.
101 P. I. Dalko and P. Sina, Angew. Chem. Int. Ed., 1999, 38, 773–777.
102 R. J. Ferrier, J. Chem. Soc. Perkin Trans., 1979, 1, 1455–1458.
103 R. Blattner, R. J. Ferrier and S. R. Haines, J. Chem. Soc. Perkin Trans, 1985,
1, 2413–2415.
104 D. H. R. Barton, J. Camara, P. Dalko, S. D. Gero, B. Quiclet-Sire and P.
Stütz, J. Org. Chem., 1989, 54, 3764–3766.
105 (a) D. H. R. Barton, S. Augy-Dorey, J. Camara, P. Dalko, J. M. Delaumény,
S. D. Gero, B. Quiclet-Sire, P. Stütz, Tetrahedron, 1990, 46, 215–230;
(b) S. Augy-Dorey, P. Dalko, S. D. Gero, B. Quiclet-Sire, J. Eustache, P. Stütz,
Tetrahedron, 1993, 49, 7997–8006.
106 (a) S. L. Bender, R. J. Budhu, J. Am. Chem. Soc., 1991, 113, 9883–9885; (b) V.
A. Estevez, G. D. Pestwich, J. Am. Chem. Soc., 1991, 113, 9885–9887.
107 S. Amano, N. Ogawa, M. Ohtsuka, S. Ogawa and N. Chida, Chem. Commun.,
1998, 1263.
108 (a) S. Adam, Tetrahedron Lett., 1988, 29, 6589–6592.
109 (a) D. H. R. Barton, S. Augy-Dorey, J. Camara, P. Dalko, J. M. Delaumény,
S. D. Géro, B. Quiclet-Sire, P. Stiitz, Tetrahedron 1990, 46, 215–220. (b) P.
Laszlo, A. Dudon, J. Carbohydr. Chem., 1992, 11, 587–593.
110 T. Iimori, H. Takahashi and S. Ikegami, Tetrahedron Lett., 1996, 37, 649–652.
111 S. K. Das, J.-M. Mallet and P. Sinay, Angew. Chem. Int. Ed., 1997, 36,
493–496.
112 M. Sollogoub, J.-M. Mallet and P. Sinaÿ, Tetrahedron Lett., 1998, 39, 3471–
3472.
113 M. Sollogoub, A. J. Pearce, A. Hirault and P. Sinaÿ, Tetrahedron, Asymm,
2000, 11, 283–294.
114 A. J. Pearce, M. Sollogoub, J.-M. Mallet and P. Sinaÿ, Eur. J. Org. Chem.,
1999, 2103–2117.
115 (a) H. Ito, Y. Motoki, T. Taguchi, Y. Hanzawa, J. Am. Chem. Soc., 1993,
115, 8835–8836; (b) Y. Hanzawa, H. Ito, T. Taguchi, Synlett, 1995, 299–305;
(c) D. J. Jenkins, A. M. Riley, B. V. L. Potter, J. Org. Chem., 1996, 61, 7719–
7726; (d) L. A. Paquette, R. M. Arbit, J. A. Funel, S. Bolshakov, Synthesis,
2002, 2105–2109; (e) A. Chénedé, P. Pothier, M. Sollogoub, A. J. Fairbanks,
P. Sinaÿ, J. Chem. Soc., Chem. Commun., 1995, 1373–1374; (f) J. J. C. Grové,
C. W. Holzapfel, D. B. G. Williams, Tetrahedron Lett., 1996, 37, 5817–5820;
(f) J. L. Chiara, S. Martinez, M. Bernabé, J. Org. Chem., 1996, 61, 6488–6488;
(g) J. M. Aurrecoechea, B. López, Tetrahedron Lett., 1998, 39, 2857–2860.
116 (a) H. Ito, Y. Motoki, T. Taguchi, Y. Hanzawa, J. Am. Chem. Soc., 1993,
115, 8835–8836; (b) Y. Hanzawa, H. Ito, T. Taguchi, Synlett, 1995, 299–305; (c)
D. J. Jenkins, A. M. Riley, B. V. L. Potter, J. Org. Chem., 1996, 61, 7719–7772.
117 D. J. Jenkins, A. M. Riley and B. V. L Potter, J. Org. Chem., 1996, 61, 7719–
7726.
118 J. M. Aurrecoechea, B. Lopez and M. Arrate, J. Org. Chem., 2000, 65, 6493–
6501.
119 J. M. Aurrecoechea, J. H. Gil and B. López, Tetrahedron, 2003, 59, 7111–7121.
120 J. M. Aurrecoechea, M. Arrate, J. H. Gil and B. López, Tetrahedron, 2003, 59,
5515–5522.
121 J. Ma, Y. Zhao, S. Ng, J. Zhang, J. Zeng, A. Than, P. Chen and X.-W. Liu,
Chem. Eur. J, 2010, 16, 4533–4540.
302 | Carbohydr. Chem., 2012, 38, 263–302

Multivalent glycoconjugates in medicinal
chemistry
José G. Fernández-Bolaños,* Inés Maya and Ana Oliete
DOI: 10.1039/9781849734769-00303
This review summarises recent advances in the design of multivalent glyco-

conjugates, such as glycodendrimers, glyconanotubes, and glyconano-
particles, and their uses in medicinal chemistry including the development of
carbohydrate-based antitumor vaccines. Literature is covered since 2007.
1 Introduction
Carbohydrates play a key role in many biochemical processes: fertilization,
immune response, inflammation, cell growth, cellular recognition and signal-
ling, viral replication, and parasitic infection, through the interaction with their
corresponding protein recognition receptors.1 Therefore, interference with these
recognition processes is a logical route for the development of new drugs.2
As the binding between an individual or monovalent carbohydrate and
an individual protein is generally weak, in the millimolar range,3 nature gets
strong and specific responses through multiple protein-carbohydrate inter-
actions, a phenomenon known as multivalency.4 Multivalency leads to
greater enhancements in binding affinity than predicted from the simple
linear additivity, and this enhancement obtained with multivalent ligands
compared to their monovalent counterparts is often referred to as the
cluster or multivalent effect.5,6
A large number of multivalent ligands have been developed to inhibit or
promote many biological processes. Most of these multivalent ligands are
glycomimetic inhibitors of lectins, in which the valency, the topology of the
epitope, the length of the spacers, and the type of the scaffold are key fea-
tures for a good cluster effect. The affinity of multivalent glycoconjugates
for lectins has been extensively described in excellent reviews.7–14
As the typical low affinity of carbohydrates for proteins is the major
drawback of their use as therapeutics, the use of multivalent carbohydrates
can be a good strategy, since they can bind much more strongly to their
targets than their monovalent counterparts.15 Affinities can be increased
several orders of magnitude due to multivalency if chelation is possible,
binding a multisite target protein.
The research on multivalent glyconjugates has experienced a dramatic
development in the last years, and an impressive number of research papers
have appeared. Only a limited number of selected references of the last five
years have been included in this review for the sake of conciseness.
2 Glycoclusters with a flexible core

A promising strategy to fight the increase in the occurrence of antibiotic-
resistant pathogens16 is to interfere with the adhesion process of the
Departamento Quı´mica Orgánica, Facultad Quı´mica, Universidad de Sevilla, 41012-Seville,

Spain. E-mail: bolanos@us.es
Carbohydr. Chem., 2012, 38, 303–337 | 303

bacteria,7 required for bacteria to establish a local site of infection
overwhelming host defenses. Pathogens are attached to specific host cells
by protein adhesins on the surface of the bacteria that adhere to com-
plementary carbohydrate receptors on the host cells. One advantage asso-
ciated with inhibiting the bacterial infection by blocking the adhesions, is
reducing the chance of resistance build-up as the pathogen is not killed,
another is to avoid problems associated to the release of toxic by-products
from dead bacteria.17
Cholera bacteria, causing cholera disease, produce the cholera toxin
(CT), a protein capable of binding simultaneously to the carbohydrate
moieties of five GM1 ganglioside molecules present on the cell surfaces of
the intestines. This binding is the key step for the internalization of the
toxin. Pieters et al. have been able to prepare using click chemistry the
multivalent conjugated versions of the GM1 oligosaccharides18 (Fig. 1).
The divalent compound 2a was an inhibitor of CT 10,000-fold stronger than
HO OH HO OH
O O
O O
HO
OH
AcHN OH
OH O HO
HOOC O Gly
O O
O
HO HO
O OH OH
HO
AcHN
HO
O O O
O
a: R = Gly 11 N N O N
H 3 H
N N
HO OH O O O
b: R = O
O O
HO
O N N O N
OH 5 5 H 3 H
N N
RHN
Gly 9
O
1
CONH
RHN
O RHN O
O O
CO2Me
RHN NH
O O
O O
RHN CONH
2a,b CO2Me
H
RHN O N O
O
O RHN O
O
RHN NH O
CONH
O
RHN O
O
O O
RHN CO2Me RHN
O O N
RHN H
4a,b
O
CONH
RHN O
O RHN
3a,b
Fig. 1 Multivalent conjugated versions of GM1 oligosaccharides.
304 | Carbohydr. Chem., 2012, 38, 303–337

the monovalent (IC50=19 mM), whereas the tetravalent and the octavalent
dendrimers 3a and 4a were 83,000-fold (IC50=0.23 nM) and 380,000-fold
(IC50=50 pM), respectively, more potent than the monovalent derivative 1.
After correction for valency, the activity of the octavalent compound is
47,500-fold stronger per sugar than the monomer.
In a similar way, that group prepared galactose-containing di-, tetra- and
octavalent dendrimers 2b, 3b, 4b with longer spacer arms.19 The penta-
saccharide of ganglioside GM1 was simplified to a b-D-galactopyranosyl
unit, the terminal monosaccharide of GM1, while the rest of the pyranose
rings were substituted by a PEG fragment. The inhibitory effects of 2b–4b
(2b, IC50=130 mM; 3b, IC50=25 mM; 4b, IC50=12 mM) were much smaller
than those observed for the pentasaccharide derivatives 2a–4a, however the
multivalency effects were strong, as deduced from the relative potency per
sugar unit: 923-fold stronger than galactose for 2b, increased to 2,400-fold
for the tetravalent stage and it remained about the same (2,500) for the
octavalent derivative.
As bacterial adhesion is the first step in the infection by pathogens,
bacterial adhesins are main targets for antibacterial therapeutics. Lectin
FimH, a specific adhesin located at the tip of type 1 fimbriae in uropatho-
genic Escherichia coli, binds to glycoproteins with terminally exposed
mannose.20 Gouin et al. have described the synthesis by Cu(I)-catalyzed
azide-alkyne cycloadditions of a tetravalent heptyl-mannosides on a pen-
taerythritol skeleton (Fig. 2). This glycocluster inhibits bacterial bladder cell
binding in the nanomolar range (12.2 nM, B6,000- and 64-fold lower than
mannose and heptyl mannoside, respectively).21 The same group has
OH
OH
O
HO
HO N N
OH O O N
OH
O 3 O
HO O
HO 4
N N
O O N
OH O O
HO 3 O
4 O
O 4
HO N N
HO
4 O
N
O O O
3 OH
OH N
O
HO N
HO
N
O O
Fig. 2 Tetravalent mannosides with a flexible core.
Carbohydr. Chem., 2012, 38, 303–337 | 305

synthesized, also by click chemistry, multimeric lactosides based on
carbohydrate scaffolds with valencies ranging from 1 to 4 and different
linker lengths. The binding affinities of the multilactosides towards galectins
showed a modest cluster effect.22
Heptasaccharidic glycodendrons as mimics of high mannose-type oligo-
saccharides (Fig. 3) could be obtained using radical addition of mercap-
toethanol to double bonds and glycosylation in a repetitive reaction
sequence. All of the glycodendrons tested as inhibitors of the mannose-
specific adhesion of E. coli proved to be better inhibitors than methyl a-D-
mannoside (MeMan). The thiahexyl-spacered glycodendrons were generally
better inhibitors than the propyl-spacered ones, probably due to increased
conformational flexibility of the spacers.23,24
In the search of new antiviral drugs able to prevent the infection
process, DC-SIGN (dendritic cell-specific ICAM-3 grabbing non-integrin),
a mannose-binding lectin present on dendritic cells and involved in the
initial stages of HIV infection, is a promising therapeutic target. DC-SIGN
is able to recognize highly mannosylated glycoproteins at the surface of
pathogens, such as HIV, dengue, Ebola and hepatitis C viruses, and
Mycobacterium tuberculosis bacterium.25,26 Therefore, the initial stages of
the infection could be prevented by the inhibition of pathogen recognition
by DC-SIGN. A tetravalent dendron containing four copies of a linear
trimannoside mimic in which the central mannose unit is replaced by a
carbocyclic diol (Fig. 4), was shown to inhibit DC-SIGN-mediated HIV
trans infection of CD4 þ T lymphocytes at low micromolar concentration.27
This dendron inhibits in a dose-dependent manner HIV-1 infection of
human cervical explants, even in presence of elevated viral load, and
displays a long lasting effect.28
Multifunctional dendrimers bearing two surface functionalities, ‘‘Janus
dendrimers’’, might be useful as drug delivery devices that can combine
tissue targeting and imaging, or be directed to a specific cell type. Boons and
co-workers29 have developed a methodology for the assembly of dendrimers
having two different functions, based on three sequential azide–alkyne
cycloadditions (Fig. 5).
This methodology is compatible with biologically important compounds
such as carbohydrates, peptides, or fluorescent labels. In this approach, a
strain-promoted azide–alkyne cycloaddition (SPAAC) between polyester
dendrons, bearing an azido and a 4-dibenzocyclooctynol moiety at the focal
point, furnished dendrimers with terminal and trimethylsilyl-protected
alkynes. The terminal alkynes were assembled with azido-modified galac-
tosyl residues by using Cu(I)-catalyzed azide–alkyne cycloadditions. TMS
deprotection and a subsequent click reaction of the resulting terminal
alkyne with azido-containing compounds gave multifunctional dendrimers,
such as that shown in Fig. 6, bearing sugars and Arg-Gly-Asp (RGD)
peptides at the periphery.29
A previously reported multivalent dendrimer with dual function,
recognition and detection, containing 16 mannose units and 2 coumarin
chromophores, exhibited 240-fold greater potency than monomeric mannose
in standard hemagglutination assays using Concanavalin A (Con A), with a
relative activity of 15 per sugar moiety when compared to mannose.30
306 | Carbohydr. Chem., 2012, 38, 303–337

OH OH
OH OH
O O
HO HO
HO
HO
OH O OH
OH OH O O
S O
O OH O OH
HO HO
HO HO O HO O
HO
S O O
O O
OH O OH O
OH OH
O O
HO O HO O
HO HO
OH OH
OH O O O O
OH HO OH O
O HO
S O OH OH O
O OH
HO O O
HO O HO HO
HO OMe HO OMe
O O
S
O
O O
O
Fig. 3 Glycodendrons as oligomannoside mimetics.
Carbohydr. Chem., 2012, 38, 303–337 | 307

OH
OH
HO O
HO
O
MeOOC O O
O
MeOOC O O
O O
O
O N3
O O
OH O O
HO O O
HO O
O
O
O O
N
H
O
Fig. 4 Linear trimannoside mimic tetravalent glycodendron.
TMS
TMS
TMS
TMS
N
1. CuAAC, N3
SPAAC N
+ N3 TMS N TMS
2. Deprotection and
CuAAC, N3
TMS
TMS TMS
TMS
N N NN
NN N N N
N N
N N
NN N N
N N N
N
N
N
N N N N N
NN N
N N
Fig. 5 Multifunctional dendrimer synthesis by three consecutive click reactions.
3 Glycoclusters with a rigid core

3.1 Glycocalix[n]arenes
Calix[n]arenes are useful cyclic scaffolds31 with well-defined conformational
properties and cavities, which can modulate shape and conformational flex-
ibility of sugar ligands. Calixarene glycoconjugates are effective in binding
important lectins such as cholera toxin,32 and the influenza A viruses.33
A series of calixarenes substituted with b-GlcNAc units linked by a thiourea
spacer has been tested for binding activity to the activation-associated
receptor CD69 (cluster of differentiation 69) on human NK (natural killer)
cells. A GlcNAc-tetrasubstituted calix[4]arene proved to be the best CD69
ligand identified to date, exhibiting stimulation of natural cytotoxicity of
human PBMC (peripheral blood mononuclear cell) in 104 and 108 M
concentrations, and inducing an increase in the spontaneous death of tumor
targets.34
308 | Carbohydr. Chem., 2012, 38, 303–337

R
N
O
N R’
R N
OO O N
N
N N
N O O N
R O
O
N O
O
N O
N O O O
R
O O O
N
R’
O N
N O
N H H
N O O N
N N O N
N O O
O
2 N N
N O O 2 H
O O N
R O O N N
N N O
N O O
N
N O
O O R’
O
R N O O
N O
N O O O N
R N O N
N
O HN NH2 N
N
R’
R HN
O O
HO OH
H H
R’ = H2N N N
R= O N N
HO O H H
O O O
OH
Fig. 6 ‘‘Janus dendrimers’’ bearing two different functions: sugars and RGD peptides.
Carbohydr. Chem., 2012, 38, 303–337 | 309

R1
R2 R2 R2
R
1 R2 HO R1 R1
OH HO HO
HO OH
OH OH
O O
O O
HO
HO HO
HO
NH NH HN
HN
S S
S HN S
NH HN
HN
cone-4Gal[4]Prop R1 = H, R2 = OH
O
O O O cone-4Lac[4]Prop R1 = O-β-Gal, R2 = H
HO OH OH
O O H
O N NH cone-4Lac[4]Prop: n = 4, R = propyl
HO HO 6Lac[6]Met: n = 6, R = methyl
OH OH 8Lac[8]Met: n = 8, R = methyl
S
OR
Fig. 7 Glycosylthioureido calix[n]arenes.
Ungaro et al. have prepared a series of calix[n]arenes (n=4, 6, and 8)

with thiourea-linked galactose or lactose moieties at the upper rim (Fig. 7).
The calix[6,8]arenes proved to be especially effective as inhibitors of
binding of three human galectins to human tumor cells in vitro, and
also effective towards a plant toxin (Viscum album agglutinin). It was shown
that inhibition is related to the conformational properties of the calix[n]-
arene scaffold, and to the shape and valency of the glycoclusters. For
example, the fixed-cone type, cone-4Lac[4]Prop (Fig. 7) could block galec-
tin-3 binding to colon cancer cells, and the flexible calixarenes 6Lac[6]Met
and 8Lac[8]Met could avoid galectin-4 binding to pancreatic carcinoma
cells.35 In a similar way, glucosylthioureido calix[6 and 8]arenes showed
interaction with plasmid DNA and agglutination with a glucose-specific
lectin such as Con A.36
A series of calix[4]arene glycoconjugates have been prepared by Vidal and
co-workers37 by coupling alkyne-derivatized calix[4]arene precursors to a
galactose azido derivative by microwave-assisted Cu(I)-catalyzed azide–
alkyne cycloadditions. The designed glycoconjugates were evaluated as
ligands for the galactose-binding PA-IL (Pseudomonas aeruginosa
agglutinin-I lectin) from the opportunistic bacterium Pseudomonas
aeruginosa, a major cause of nosocomial lung infections for immuno-
compromised patients.38 The interaction with PA-IL lectin was shown to be
strongly dependant on both the valence and the topology, with tetravalent
310 | Carbohydr. Chem., 2012, 38, 303–337

HO OH 4
O O
O O
HO Conformation Kd [nM]
O N
OH
cone 420 ± 160
N N partial cone 200 ± 5
HO OH
1,3-alternate 176 ± 6
O OH
O O
HO
O N Kd 150,000 ± 33,000 nM
OH
N N
Fig. 8 Three topologically isomeric tetravalent calix[4]arenes.
galactose ligands binding most strongly. Nanomolar affinities were mea-

sured by isothermal titration microcalorimetry (ITC).
Three topologically isomeric tetravalent conjugates were compared,
the one locked in 1,3-alternate conformations, topology 2:2 (Fig. 8) was the
most active, with an 800-fold enhancement of affinity compared to the
monovalent derivative. The one in the cone conformation, topology 4:0,
showed only a 360-fold enhancement. Atomic force microscopy has
revealed that PA-IL lectin and the 1,3-alternate glycocluster create
monodimensional filaments according to an aggregative chelate binding
mode.39
A recent report by Cecioni et al.40 describes the synthesis of a library of
optimized tetravalent glycoclusters (Fig. 9) using triethylene glycol spacers
(a) and three structurally modified analogues that incorporate one or two
amide moieties and/or a phenyl ring (b–d), in order to study the influence of
the flexibility of the linker arm on the interaction of the prepared gly-
coclusters with PA-IL lectin. The azido-functionalized precursors contain-
ing per-O-acetylated galactose were coupled with propargyl acetate to get
the monovalent references and with tetra-propargylated core scaffolds
(Zn-porphyrin A, methyl a-D-glucopyranoside B, cone-calix[4]arenes C,
partial cone conformer D, and 1,3-alternate conformer E) to afford acety-
lated glycoclusters. The coupling was carried out by CuAAC, under
microwave irradiation.
Monovalent p-amidophenyl-based ligand 6 (Fig. 9) showed a dissociation
constant value (KD) with PA-IL, determined by ITC, of 5.8 mM, which is an
unprecedented affinity for a monovalent glycomimetic ligand toward this
lectin, and much higher affinity than that shown by the analogues with more
flexible spacers analysed in the article (107–181 mM). The binding of
the prepared glycoclusters was also analysed by ITC, and 5bE with a 1,
3-alternate calix[4]arene as the core and one amido moiety in the spacer
showed the best affinity (KD value of 90 nM),40 whereas for the more flexible
5aE the KD value was 176 nM.37
Other series of tetravalent calix[4]arene glycoclusters incorporating
b-lactosyl residues to the upper or to the lower rims of the macrocycle has
been prepared by Field et al. by using click chemistry methodology.41 These
Carbohydr. Chem., 2012, 38, 303–337 | 311

OH HO Linkers
OH HO
HO
O O O
HO OH O O
O a O N3
HO O
Li
nk er 9 atoms
er nk
N N Li
N N O H
N N O N
b O N3
O
9 atoms
1
O
O H
N O N
N N N c N N3
r N N L in H
ke ke O
Lin r
OH O O 9 atoms
HO O HO O
O
d O
OH OH
HO
OH 5 O
N3
N
HO H
8 atoms
O
O
O OMe
O O O
N N O
O Zn O O O
N N
B C
O
O
A
O O
HO OH
O O O
O O O
HO O
O N N
OH
N OH
N
H
6 D E
Fig. 9 Structure of the four linker arms (a–d) and five cores (A–E), coupled to give tetravalent
glycoclusters 5.
compounds showed trypanocidal activity against T. cruzi, with one of them

showing similar activity to benznidazole, an established anti-trypanosomal
drug.
3.2 Glycofullerenes
Viruses such as retrovirus HIV are spherical particles decorated with
envelope glycoproteins.42 These glycoproteins are responsible to interact
with cell surface receptors in a multivalent mode and activate a cascade of
processes that lead to cell attachment and entry of the virus into the cells.
Therefore, they are important targets for both therapeutic and prophylactic
approaches.43 [60]Fullerenes covered by carbohydrates (‘‘sugar balls’’)
could mimic the virus surface and could interact in a multivalent manner
with cell surface receptors. In this context, fullerene hexakis-adducts bear-
ing 12 or 24 peripheral carbohydrate moieties were prepared using click-
chemistry methodology, thus obtaining a spherical platform for a globular
312 | Carbohydr. Chem., 2012, 38, 303–337

R
R
R
N N
N N N
N
N N
N
R
R
N
O N N
N O
O
O O
N
N O
O
O
O O
R
O
N N O N
O
O N N
N
O O
R O
O O N
OO N
O O N
O
R
N N
N
R N
N N
N N N
N
N R
N
OH
OH
O
OH HO O
OH HO
H
O N
HO O O O
HO OH
OH O
O O O
HO
HO H
N
O O
Fig. 10 Fullerene hexakis-adducts bearing 12 or 24 peripheral carbohydrate moieties.
multivalent presentation of ligands.44,45 The mannosyl derivatives (Fig. 10)

were recognised by Con A in a multivalent manner, showing that fullerene is
an efficient platform for the globular presentation of carbohydrates for
protein–carbohydrates interactions. The DG values of the binding process
for the glycofullerene with 12 mannose units is apparently more favourable
than with 24, due to the higher entropic penalty for the binding process with
the latter.
4 Multivalent glycopeptides in immunotherapy

Cancer cells show aberrant glycosylation of cell surface glycoproteins, with
many glycosyl epitopes constituting tumor-associated antigens. Although it
Carbohydr. Chem., 2012, 38, 303–337 | 313

remains unclear whether aberrant glycosylation is a result or a cause of
cancer, this process is a key event in induction of invasion and metastasis.46
In epithelial cancers, mucin glycoproteins over-express O-linked glycans
called tumor-associated carbohydrate antigens (TACAs) such as Tn
(GalNAca1-O-Ser/Thr), TF (Thomsen-Friedenreich, Galb1-3GalNAca1-
O-Ser/Thr) and the corresponding sialylated derivatives STn and STF,
which are considered a promising target for immunotherapy.47,48 Glycoli-
pids such as gangliosides are also overproduced on the surface of various
cancer types.49
The immune responses towards glycoconjugates rely on T-cell triggering
mechanisms depending on the nature of the carbohydrate antigens. How-
ever these carbohydrate antigens exhibit poor immunogenicity as carbo-
hydrates alone are usually T-cell independent antigens.50 Carbohydrate
antigens activate independent T-cell responses which create short-lived
immunoglobulin G (IgG) antibodies. Therefore, they cannot be employed
as vaccines alone. In the last few years there has been an impressive
development of the design of synthetic multi-component MUC1-based
glycopeptide vaccines.51–59 These vaccines rely on mimicking the natural
presentation of TACAs, typically found in clusters, for generating strong
antibody responses and immunogenicity.60,61 In this context, glycoden-
drimers have shown to be useful in vaccine preparation, due to their
immunostimulating properties.62
Kunz et al.,63 according to the principle of multiple antigen presentation,
prepared a tetrameric vaccine candidate (Fig. 11) consisting of a di-lysyl-
lysine core and four branches formed by a T cell epitope from tetanus
toxoid and a sialyl Tn-glycododecapeptide of MUC1. In a similar way, the
same group prepared by solid-phase procedures tetrameric and octameric
MUC1 and MUC4 Tn-glycopeptide vaccine candidates.64 They have also
described the preparation of efficient fully synthetic antitumor vaccines by
the combination of tumor-associated mucin glycopeptide antigens with
lipopeptide Toll-like receptor 2 ligands, able to induce immune reactions
in mice.65
Other antitumor vaccine candidates have been developed, consisting
of mono-, di-, and tetravalent MUC1 glycopeptides conjugated by click
OH CO2H
HO OH
O
AcHN O
HO
HO
O
HO
AcHN
O O NHYSYFPSVC(O)NH O O
R= Ac-GVTSAPDTRPAPC(O)NH O O
O O
NHR
H O
N
RHN OH
O
NHR
HN
RHN
O
Fig. 11 Tetravalent sialyl-Tn-MUC1-Lys2-Lys1 glycodendron vaccine.
314 | Carbohydr. Chem., 2012, 38, 303–337

HO OH
HO O
O OH OH
AcNH O HO
HO2C O O O N
R= O
O OH N N O
OH
HO HO
H
AcHN OH R N Th Epitope
HO
H O
N
R N Th Epitope
H
N
R H
Fig. 12 Synthetic GM2 neoglycopeptide.
chemistry with immunostimulating toll-like receptor2 (TLR2) lipopeptide

ligands. Oligovalent glycopeptide–lipopeptide conjugates are considered
more immunogenic than their monovalent analogues.66
A 40 amino acid MUC1 glycopeptide bearing 10 copies of the Tn antigen
has been prepared by Payne and co-workers using the direct aminolysis
ligation of a peptide thioester with a deprotected MUC1 glycopeptide.67
The same group described the synthesis of MUC1-lipopeptide vaccine
candidates, consisting of MUC1 glycopeptides, bearing multiple copies of
Tn and TF tumor-associated carbohydrate antigens tethered to the lipo-
peptide immunoadjuvant Palmitoyl3CysSer.68
A neoglycopeptide immunogen displaying one or two copies of the tet-
rasaccharidic moiety of GM2 ganglioside (Fig. 12), an important target
for specific anticancer immunotherapy, has been prepared by ligation of
alkyne-functionalized GM2 with an azido CD4 þ T cell epitope peptide.
These synthetic glycopeptides can induce human tumor cell-specific anti-
bodies in mice after immunization. It is noteworthy that the monovalent
presentation of GM2 peptide induced antimelanoma antibodies, but failed
the multivalent form. This is considered the first report on the induction of
human tumor cell-specific antibodies after immunization with a synthetic
glycopeptide.69
Danishefsky et al. have developed glycoconjugated vaccines that are
useful for the treatment of cancer and effective at inducing antibodies
against all carbohydrate antigens present on the glycoconjugates. For
example, vaccine shown in Fig. 13 was prepared by conjugation of a pen-
tavalent glycopeptide, containing five prostate and breast cancer associated
carbohydrate antigens (Globo-H, GM2, sialyl Tn, Tn and TF), to
maleimide-activated carrier protein KLH (Keyhole Limpet Hemocyanin).
These vaccines were able to induce in mice IgM and IgG (immunoglobulins
M and G) against each one of those five carbohydrate antigens.70,71
In the design of carbohydrate-based vaccines the choice of the right
template for multivalent carbohydrate presentation is a key step, because
very flexible scaffolds will decrease effective clustering of the antigen.
Danishefsky and co-workers have synthesized cyclic peptide scaffolds72
(Fig. 14) upon which the clustered tumor-associated antigens Tn and STn
Carbohydr. Chem., 2012, 38, 303–337 | 315

OH COOH
HO OH
O
HO OH HO HO AcHN O
OH OH
HO HO HO OH
O O O STn
HO O O O O
HO HO Tn
O NHAc HO
O OH AcHN AcHN
OH
O O O O
OH HO HO
O O
OH O
HO OH
Globo-H OH
4
O O
H H H H
N N N N
AcHN N N R
H H 3
O O O
HO 4
316 | Carbohydr. Chem., 2012, 38, 303–337

OH
O
HO O
OH O
GM2 NHAc HO
O
O
HOOC O HO HO TF
O O
HO HO AcHN O
O OH OH
HO HO
OH O
AcHN O HO
HO
O HO
O OH
R= N
KLH
S
O
O
Fig. 13 Pentavalent glycopeptide, containing five carbohydrate antigens (Globo-H, GM2, sialyl Tn, Tn and TF) conjugated to carrier protein KLH.
HO HO
OH OH
HO HO HO OH HO OH
OH AcHN
OH OH AcHN OH
OH CO2H HO CO2H
OH HO O O
OH OH STn
AcHN AcHN
AcHN AcHN
CO2H
HO HO O HO O
HO O CO2H HO HO
O CO2H O O
CO2H
Tn
HO HO HO
HO O O
HO HO OH HO
HO O O HO O
HO HO OH
HO O
HO AcNH HO O AcNH
O
O O O O O O
AcNH
AcNH O
AcNH AcNH
AcNH O AcNH
O O
O O
HO2C HO2C
HO2C HO2C
O NH O NH
NH NH
O HO2C O HO2C
HO2C NH NH HO2C NH NH
O O O O
N A N N A N
P F C p P F C p
N A N N A N
p Y Y P p Y Y P
t SStBu
SS Bu
Fig. 14 Cyclic peptide scaffolds as vaccine precursors.
Carbohydr. Chem., 2012, 38, 303–337 | 317

would be displayed in a well-defined orientation. A cross-linker has
been introduced by a ring-closing metathesis-mediated route to control the
distance between glycans attached to the peptide platform. This type of
cyclic peptide scaffolds has been employed by the same group in the
development of an HIV vaccine.73,74 In this context, Wang et al. have
developed a strategy for the synthesis of highly functionalized oligomannose
dendrons in the search of a HIV vaccine,75 and Kabanova et al. prepared
polyamidoamine (PAMAM) dendrons displaying oligomannose clusters of
HIV-related antigens, able to induce IgG antibodies.76 Carbohydrate based
vaccines against bacteria and parasites have also been developed.77–79
5 Glycoconjugates on tubular scaffolds

5.1 Glycocarbon nanotubes
Single-walled carbon nanotubes (SWNTs) can be functionalized with car-
bohydrates, enabling effective interactions with cells and other biological
species that are not available to the free bioactive molecules. For example,
SWNTs functionalized with mannose or galactose (Man-SWNT or Gal-
SWNT) (Fig. 15) aggregate with Bacillus anthracis and Bacillus anthracis
spores in the presence of divalent cations to result in significant reduction in
colony forming units (CFU).80,81 The effectiveness in the CFU reduction of
B. subtilis spores by Man-SWNT was in the order of Mg2 þ oCa2 þ oBa2 þ .
It was surprising that the divalent or tetravalent sugar dendron-function-
alized SWNTs (Man2-SWNT, Gal2-SWNT, Man4-SWNT, and Gal4-
SWNT) were found to be generally ineffective or less effective in binding and
aggregating the Bacillus spores in the presence of Ca2 þ .81 Mannose and
galactose dendron-functionalized SWNTs were also evaluated in binding
assays with pathogenic E. coli.82
5.2 Tubular supramolecular architecture

Columnar supramolecular polymers (Fig. 16) have shown to be effective
polyvalent scaffolds for detecting and binding two strains of Escherichia coli
(ORN178 and ORN208).
The self-assembly into columnar polymers of disc-shaped molecules such
as shown in Fig. 17, decorated or not with sugars, enhances the potency of
the monomers significantly more than expected on the basis of the multi-
valent effect alone. It was also observed that an effective decreasing of the
amount of mannose component of these polymers actually enhanced bac-
terial aggregation, due to a reduction of steric crowding.83
6 Glycodendrimer chips
Carbohydrates can be displayed on microarray chips to be screened by large
series of carbohydrate binding proteins and pathogens.84–86 These studies
are useful for the design of new drugs using only a tiny amount of material.
Pieters et al.87 have displayed glycodendrimers on a microarray surface in
such a way that the number of carbohydrates on each spot is the same while
changing the valency (Fig. 18). Spots containing from monovalent a-D-
mannopyranoside compounds to octavalent dendrimers were evaluated
with fluorescent Con A and GNA (Galanthus nivalis agglutinin) lectins, and
318 | Carbohydr. Chem., 2012, 38, 303–337

HO OH O
O
HO O
N
OH H
HO OH
O NH
HO O
OH O O
HO OH
O NH O O
HO O
OH
O N
H
HO OH
O NH
HO O
OH O O
HO OH
O NH O
HO O
OH O
O
O
N
H
HO OH O O
O
HO O
NH O
OH
HO OH O O
O
HO O
NH
OH
Fig. 15 Galactose-functionalized SWNTs.
Mannose-functionalized monomer
Unfunctionalized monomer Multivalent

supramolecular polymer
Fig. 16 Multivalent mannose-functionalized and unfunctionalized columnar supramolecular

polymers.
Carbohydr. Chem., 2012, 38, 303–337 | 319

O NH N
N HN O
O
N
N
H
O NH
N
NH O
N
N
HN
OH
OH
O N O
HO
HO N N 5
O
Fig. 17 Mannose functionalized disc-shaped molecules that self-assemble into columnar

supramolecular structure.
the binding was observable in real time. In a single experiment it was

possible to observe the multivalency enhancement, small for ConA, in
agreement with its widely spaced binding sites, and large for GNA,
according to its more closely spaced binding sites.
T. cruzi displays host-dependant cell surface diversity dominated by
mucin O-linked glycoproteins, and sialylation of non-reducing terminal
galactose residues of these mucins has been related to evasion of the
immune response.88 Field et al. have investigated the role of sialic acid
recognition in host infection by Trypanosoma cruzi using surface plasmon
resonance imaging (SPRi) as a real-time analysis of protein interactions
with carbohydrate microarrays.89 SPRi of an array of 40 sialylated and
unsialylated glycans determined the binding preference of hSiglec 7 (human
sialic acid binding Ig-like lectins) for a-2,8-linked disialic acid, whereas
murine siglec E showed preference for terminal sialic acid.90 The same
technique was used for screening ligands of the ricin surrogate RCA120.
A non-natural 6-modified galactoside exhibited up to a 3–4 fold enhanced
RCA120 binding compared to the parent galactoside.91
7 Glyconanoparticles
Carbohydrate-based nanoparticles or glyconanoparticles (GNPs) are
structures that mimic carbohydrate presentation of glycocalyx. They consist
of a surface covered with multiple carbohydrate residues, which enhances
320 | Carbohydr. Chem., 2012, 38, 303–337

HO OH
O
HO
2 HO N
R = N
O N
monovalent
1
R = HN O NH2
3
2
RO
divalent CONH
2 2
RO RO O
2
RO O
1 1
COR COR
2 NH
RO O
2
RO O
CONH 1
COR
2
tetravalent RO
2 H
RO N O
O
CONH 2 O
RO
O NH O
2
RO 1
COR
2 2 O
RO OR
octavalent O
O 2
RO N
CONH
H
2
RO
2
OR
Fig. 18 Microarrays surfaces with identical mannose contents and different valencies.
Carbohydr. Chem., 2012, 38, 303–337 | 321

O
Carbohydrates
O
Cell targets
HN
O
O
S S
S
O S H
N
Au S GRKKRRQRRR
S
O Peptides
S S
HOOC
S
N
NH
HO Gd
3+ N
N
COOH
N
RNA
O COOH COOH
Contrast agents
O
Fluorescent Dye
Fig. 19 Multifunctional glyconanoparticles and possible molecules to be attached on the

surface: fluorescent probes, peptides, DNA/RNA, metal chelates and cell targets, with linkers
of different nature and length.
affinity binding with the corresponding receptor, and an inorganic and

nano-sized core which provides useful optical, electronic and magnetic
properties depending on the chosen material.92
Some features of this model are its globular shape, controllable nanometer
size, multivalent display of carbohydrates and easy achievement of multi-
functionality on the inorganic surface. The modification of the density of
sugar residues and the nature of the linker modulates accessibility to the
ligands and their behavior during the molecular recognition events (Fig. 19).93
Nanoparticles (NPs) have many applications in clinical medicine as drug
carriers94,95 or biosensors.96 Among NPs, gold and magnetic NPs are pro-
mising tools for medical imaging, diagnosis and therapy.97–99 Carbohydrate
coating confers favorable properties to metallic nanostructures for bio-
medical applications such as stability and solubility in biological media,
biocompatibility, non-toxicity and improved specificity. Carbohydrates
facilitate distribution and penetration of glyconanoparticles into cells and
thus, GNPs are attracting attention as anti-HIV agents,100,101 anti-cancer
vaccines102 and magnetic probes in cellular labeling and Magnetic Reso-
nance Imaging (MRI).92,103
Recent biomedical applications of glyco-NPs in protein recognition,
inhibition of the adhesion of virus, bacteria and toxins, cellular targeting
and as probes in fluorescence and MRI will be described in this section.
We specially highlight examples of metallic nanoparticles in which the
sugar ligands, covalently linked to the metallic nucleus, play a key role in
biological recognition processes and not the ones where carbohydrates are
used as stabilizing agents.104 Examples are divided into three subsections
according to the NP core composition: gold NPs, magnetic NPs and
quantum dots.
322 | Carbohydr. Chem., 2012, 38, 303–337

7.1 Gold glyconanoparticles
7.1.1 Carbohydrate-carbohydrate interaction
Carbohydrate-carbohydrate interactions may be considered as a first step in
cell-cell recognition processes. They are dependent on divalent cations and
very specific. However, the low affinity between carbohydrates needs to be
compensated with multiple binding events. Multivalent presentation of
carbohydrates is therefore a requirement for potential glycocalyx-like
surface models.
Gold NPs have been successfully employed in the study of weak carbo-
hydrate-carbohydrate interactions since first reports by Penadés and co-
workers.105,106 Gold NPs coated with the trisaccharide Lewis X (Lex) were
used to study the calcium-dependent Lex self-aggregation, demonstrating
the self-recognition capability of this antigen.105
Similarly, Kamerling et al. studied carbohydrate-mediated self-recogni-
tion of marine sponges. Using gold NPs coated with a sulfated disaccharide
and with a pyruvated trisaccharide naturally occurring in the extracellular
surface of marine sponge cells, and studying by AFM (Atomic Force
Microscopy)107 its Ca2þ-dependent self-recognition, they concluded that the
self-interaction between the sulfated disaccharide fragments is stronger than
that between the pyruvated trisaccharides. They have also employed NMR
spectroscopy108 [transfer-nuclear Overhauser effect spectroscopy (TR-
NOESY) and diffusion ordered (DOSY)-NMR] to obtain structural
information related to these interactions.
Russell et al. prepared 16 nm gold nanoparticles coated with lactose
through thiol-ended chains of different lengths and studied the calcium-
induced aggregation by TEM (Transmission Electron Microscopy) and
UV-visible spectroscopy.109 It was shown that the length and nature of the
linker had an important effect on the metal ion-mediated carbohydrate-
carbohydrate interaction of the lactose nanoparticles.
Aminooxy-functionalized Au nanoparticles lead to glyconanoparticles
when reacting with glycosphingolipid (GSL) aldehydes, derived by ozono-
lysis of natural glycosphingolipid expressed in mouse melanoma B16
cells.110 The technology named ‘‘glycoblotting’’ (Fig. 20), consisting in
glycan enrichment based on chemical reaction, has led to the evaluation
by SPR of the interaction between GM3-funtionalized GNPs and self-
assembled monolayer (SAM) of Gg3Cer (gangliotriaosylceramide).
7.1.2 Carbohydrate-protein interactions

Studies on carbohydrate–protein interactions using gold GNPs111 are
mostly based on colorimetric methods detecting changes in the surface
plasmon resonance (SPR) of Au nanoparticles. Gold GNPs have been
developed into a colorimetric biosensor for studying these interactions.112
The specific recognition by proteins of carbohydrate covering gold GNPs
leads to aggregation of nanoparticles causing a color change associated with
a shift in the surface plasmon absorption-band of the metal cluster, and this
can be monitored using UV-visible spectrophotometry.
Kataoka et al. studied the specific and reversible interaction between
lactoside-coated gold NPs and Ricinus communis agglutinin (RCA120).113
Carbohydr. Chem., 2012, 38, 303–337 | 323

HO O
OH OH OH
HOOC OH
O HN
O 7
HO O O O O GM3
AcHN OH HO
HO OH 6
OH
Ozonolysis
HO O
OH OH OH GM3 aldehyde
HOOC OH
O HN
O 7
HO O O O O
AcHN OH HO O O
HO OH
OH
GSLs-blotting
O HO
HO S
S GM3 aldehyde
Au N O S Au
H2NO S
SPR-based functional analysis
Gg3Cer-functionalized SAM
HO
GM3-functionalized GNP
S S S OH
S
Au
Fig. 20 Glycoblotting methodology and interaction between GM3-functionalized GNPs and

Gg3Cer-functionalized SAM.
It was shown that the degree of aggregation is dependent on lectin

concentration and on nanoparticle ligand density. Other groups have also
worked on colorimetric assays based on aggregation of gold nanoparticles
because of their simplicity and versatility. For example, Russell et al.
synthesized carbohydrate-stabilized gold nanoparticles to detect cholera
toxin114 and RCA120,115 showing that the degree of aggregation is
dependent on chain length in addition to the surface ligand density.
Jensen and co-workers tested oxime- and oxyamine-linked (oligo)gluco-
side-GNPs with Con A, the most broadly used lectin in carbohydrate–protein
324 | Carbohydr. Chem., 2012, 38, 303–337

interaction studies.116 The ring-closed tautomer of N-glucosyl oximes is
recognized by the lectin, whereas the reduced, open chain form, displayed a
complete lack of aggregation. In a recent paper Thygesen and Jensen review
methods for preparing glyconanoparticles based on chemoselective formation
of thiol-ended glycoconjugate intermediates and those ones based on
chemoselective reactions between reducing carbohydrates and reactive
functional groups on polymer coated nanoparticles.117
Interaction between galactose functionalized AuNPs and PA-IL lectin
from Pseudomonas aeruginosa has been evaluated by hemagglutination
inhibition assay (HIA), SPR and isothermal titration calorimetry assays,
being Gal-AuNPS the most efficient binding partner for this lectin to date,
with a Kd of 50 nM.118
Recognition studies of maltose- and maltotriose-NPs of different ligand
density with glucoamylase concluded that lower carbohydrate density
facilitates accessibility of the sugars to the enzyme.119 Penadés and co-
workers have synthesized lactoside-functionalized NPs and studied their
interaction with Viscum album agglutinin and their hydrolysis with E. coli
b-galactosidase.119 These results proved the influence of the density and
presentation of sugar residues toward protein recognition and enzymatic
hydrolysis. Thus, selection of correct ligand density is an important factor in
order to achieve the desired affinity while escaping degradation by glyco-
sidases. The activity of polysaccharide-degrading enzymes related to
pathophysiological alterations in human diseases has also been tested using
polysaccharide-GNPs.120
Biosensing methods other than colorimetric ones have been developed.
Biotinylated carbohydrate-based copolymers containing b-GlcNAc or
a-Man residues were conjugated to the surface of gold nanoparticles and
immobilized on an avidin-coated DOTLabs sensor chip (Diffraction Optic
Technology) to study the specific recognition of carbohydrates with lectins
Con A or wheat germ agglutinin (WGA),121 the first one recognizing
a-Man, and the second one recognizing b-GlcNAc residues.
Interaction between sialic acid-coated gold nanoparticles and Alzheimer’s
amyloid-beta (Ab) peptides has been electrochemically tested.122 Yan and
co-workers developed a fluorescent competition binding assay method to
measure the binding affinity of GNPs-lectin interaction, showing that
carbohydrate ligands conjugated to AuNPs exhibited affinities up to five
orders of magnitude higher than those of the corresponding monomeric
ligands with lectins.123 Cyanovirin-N, a cyanobacterial lectin which exhibits
inhibitory activity against a number of viruses, including HIV, has been
recently evaluated by a fluorescence-based competition assay.124 Further-
more, a method for protein detection based on surface energy transfer
(SET) has recently been reported to study the interaction between
fluorescent labeled FITC-Con A (fluorescein isothiocyanate-Con A) and
mannose-functionalized-dendrimer covering gold nanoparticles.125
Among other analytical applications, gold GNPs have been successfully
employed to isolate and purificate lectins,126 and for the quantitative
determination by SPR of small molecular weight analytes like galactose,
using lactosyl-(polyethylene glycol)-gold NPs adsorbed onto Ricinus
communis agglutinin (RCA120).127
Carbohydr. Chem., 2012, 38, 303–337 | 325

7.1.3 Gold GNPs as non-viral gene delivery agents
Non-viral vectors are attracting much attention in gene therapy in order to
overcome the safety problems of their viral counterpart.128 Gold GNPs may
be used as non-viral gene delivery agents as Glc- and Gal-gold nanoparticles
were highly efficient DNA binders. Interaction between GNPs and DNA
showed binding affinity dependence on the sugar ligand.129
7.1.4 Gold GNPs as antiadhesion agents

Many pathogenic infections start through multivalent carbohydrate-
dependent adhesion of microorganisms to host cells. Carbohydrate-func-
tionalized nanomaterials could act as antiadhesion agents avoiding the
adhesion of microbes to host cells and thus, becoming an alternative to
antibiotic strategy. Manea et al.130 have assembled thiolated synthetic
analogues of type A Neisseria meningitides antigens on the surface of gold
nanoparticles to inhibit the binding of this bacterium to specific mouse
polyclonal antibodies.
Gold Gal- and Glc-NPs were employed to evaluate their ability to displace
HIV envelope glycoproteins gp120 from plate-bound galactosylceramides.131
Promising results have also been achieved by Penadés and co-workers
when using gold GNPs to study the multivalent interactions with gp120. They
prepared a set of hybrid high-manno-NPs to mimic the presentation of
oligomannosides on the HIV envelope protein gp120 and have studied them
as inhibitors of DC-SIGN–gp120 interactions.100 GNPs containing the dis-
accharide Mana1-2Mana (Fig. 21) showed more than 20,000-fold increased
activity of inhibition compared to the monomeric disaccharide.
These high mannose GNPs, mimicking the cluster presentation of
glycoprotein gp120 on the virus surface,100 were tested in cell-based
experiments for evaluating their anti-HIV potential as inhibitors of
DC-SIGN-mediated HIV-1 trans-infection of human T cells.101 Low per-
centages of infection compared to untreated control were found; therefore
HO2C
OH O
OH
HO O O
HO
OH 5
O S
O
HO 5
HO Au
H
O N O S
O
5 5
O
OH OH
OH
O HO
HO O
HO HO
OH O
O OH
HO O
S S
HO
O Au
O S
N N
H H 4 4
Fig. 21 Hybrid Mana1-2Mana glyconanoparticles.
326 | Carbohydr. Chem., 2012, 38, 303–337

these GNPs may function as anti-adhesive barrier at an early stage of HIV
infection.
Affinity information on the molecular recognition of synthetic linear
and branched-oligomannosides with the anti-HIV-1 antibody 2G12 has
been reported using NMR techniques, showing that repetition of binding
motifs or increased complexity of the oligomannosides do not necessarily
involve an enhancement of affinity.132 AuNPs coated with a linear tetra-
mannoside conjugate inhibit 2G12-gp120 binding with IC50 values in the
micromolar range, and could block the 2G12-mediated neutralization of a
HIV infection, under conditions similar to the ones in which normal serum
prevents infection.133 It has been described that glucose used as a second
ligand in folate coated-gold nanoparticles enhanced cell recognition of
cancer cells as a result of multivalent interactions between both ligands and
cells.134
7.1.5 Gold GNPs as antitumor agents

The specific anti-metastatic effect of lactose-GNPs on development of lung
tumor foci in mice has been demonstrated.135 Mice were injected with
melanoma cells (B16F10) pre-incubated with Lac-GNPs and after three
weeks, the lungs were analyzed and compared to the control ones, showing
an inhibition up to 70% of metastasis in lung cells of mice. Heparin coated
AuNPs exhibited effective inhibition of basic fibroblast growth factor 2
(FGF-2)-induced angiogenesis in a mouse model. Naked Au administrated
at the same concentration resulted lethal to the animals, demonstrating that
carbohydrates impart biocompatibility to these particles.136
7.1.6 Gold GNPs as tool for pathogen and toxin detection

Pathogens often have carbohydrate-binding proteins to recognize and
invade host organisms or to deliver toxins. Therefore, GNPs could be used
as a new methodology for rapid and sensitive pathogen and toxin detection.
For example, globotriose-functionalized gold NPs have been designed for
detection and inhibition of verotoxin, a Shiga-like toxin.137 These AuNPs
presenting carbohydrate analogues to the glycolipids on cell surface can
effectively compete for these toxins, being the inhibition dependent on the
structure and density of the glycans.138
Russell and co-workers reported a rapid colorimetric bioassay based on
SPR for detection and quantification of cholera toxin using carbohydrate-
stabilized gold nanoparticles.114 Glyconanoparticles were also used by
Field and co-workers to elucidate structural and mechanistical information
of the role of sialic acid recognition in the host infection by Trypanosoma
cruzi. These nanoparticles designed as T. cruzi mucin glycan mimetics were
used in lectin binding assays and the results highlighted the importance of
the branched structures of mucin glycans in biological recognition. In
addition, sialic acid tethered to gold nanoparticles was used to detect siglec
lectin specificity.89
7.1.7 Gold GNPs for molecular and cellular imaging

There is an increasing demand for new probes in the field of in vivo optical
imaging. Quantum dots (QDs), magnetic and gold nanoparticles have
Carbohydr. Chem., 2012, 38, 303–337 | 327

demonstrated to be sensitive and target-specific imaging. Herein we report
some examples where carbohydrate-functionalized gold nanoparticles have
also been applied for in vivo imaging.
Multifunctional gold GNPs coated with hyaluronic acid (HA) residues
labeled with a fluorescence dye have been used for in vivo imaging of arthritic
inflammation and human ovarian carcinoma tumor (OVCAR-3) in mice.139
Ex vivo examination of tissues revealed large amounts of HA-AuNPs accu-
mulated in tumor sites. A strategy to confer magnetic properties to noble-
metal GNPs consists in the functionalization of the metal surface. Hybrid gold
nanoparticles coated with sugar conjugates and Gd(III) complexes have been
prepared as paramagnetic probes for Magnetic Resonance Imaging (MRI)
and were used for in vivo imaging of glioma in mouse brain.140 The behavior of
these GNPS as in vivo contrast agents is dependent of the nature of the sugar
and the relative position of the sugar with respect to the gadolinium ion.
7.2 Magnetic glyconanoparticles

Magnetic glyconanoparticles (MGNPs) are materials with potential appli-
cation in the field of clinical diagnosis and therapeutics and have mainly
been used as contrast agents for MRI.
7.2.1 Lectin interaction and bacterial detection with MGNPs

Carbohydrates attached to magnetic NPs could be used in the recognition
of lectins such as Con A. For example, fluorescein isocyanate-labeled Con A
recognizes Man-MGNPs and could be separated from the medium by a
magnetic field.141
MGNPs could also serve as a method for pathogen detection and
decontamination, such as E.coli ORN178 containing mannose binding
proteins on surface. Silica-coated iron oxide magnetite nanoparticles,
functionalized with mannose or galactose residues by click chemistry or
peptide coupling were used by the group of Huang142 to detect E. coli
strains with a detection limit of 104 cells/mL. Taking advantage of their
magnetic properties, up to 88% of the bacteria were removed from the
medium (Fig. 22). Furthermore, based on the response patterns three E. coli
strains could be differentiated. Detection of pathogenic Streptococcus suis
bacteria was achieved using magnetic glyconanoparticles bearing biotiny-
lated sugar and an ATP detection system.143
7.2.2 Cell binding with MGNPs

Gold-iron oxide superparamagnetic glyconanoparticles were first synthe-
sized by reduction in water a mixture of gold and iron salts in the presence
of thiol-ended glycoconjugates.144 The interaction of these Fe/Au nano-
particles with human fibroblast cell lines was tested, showing that these cells
recognize selectively the sugars on the nanoparticle surface.145
Several hyaluronic acid-coated magnetic nanoparticles have been reported
as target-specific MR imaging probes in cancer cells146 and macrophages.147
Recently, El-Boubbou141 prepared MGNPs with different monosaccharides
as a nanosensor system to detect and differentiate nine types of cancer cells
using a statistic method based on different response patterns.
328 | Carbohydr. Chem., 2012, 38, 303–337

A
O
O O
O
O
Magnetic glyconanoparticles
O Fe3O4 (MG NPs)
O
O
O
O O
O
MAGNET
1) MGNP Detection Capture
Quantification efficiency
2) Magnetic
separation
E. coli sample
Fig. 22 Escherichia coli detection and decontamination with MGNPs. A) Silica coated
magnetic glyconanoparticles. B) Pathogen detection.
Some recent examples of multimodal magnetic immuno-glyconano-

particles designed for immunolabeling and imaging of cells have also been
reported.148,149 In these nanoparticles, carbohydrates confer them bio-
compatibility, stability and resistance to nonspecific adsorption.
7.2.3 In vivo applications of MGNPs

Diverse carbohydrate-functionalized magnetic nanoparticles, synthesized
by chemical functionalization of previously prepared iron oxide nano-
particles, were design for in vivo inflammation detection. sLeX-MGNP
resulted to be a highly sensitive and selective T2 contrast agent for direct
detection of E-/P-selectins, endothelial markers in acute inflammation.150
7.3 Glyco quantum dots

NPs can be improved by modifying their surfaces by attaching a fluorescent
dye or by modifying the nature of the core with semiconductor nanocrystals
like CdSe, CdTe or InP (quantum dots, QDs). A combination of the two
methodologies could lead to a bimodal imaging probe.151 Chemical and
physical properties of quantum dots (size-tunable light emission, bright
luminescence, broad absorption spectra and resistance to photobleaching)
have made this material popular as bioluminescent probes for live-cell and
in vivo animal imaging.152
7.3.1 Protein recognition

QDs, as well as silver and magnetic GNPs, functionalized with dextran have
been used to study the interaction with lectins such as Con A.153 The specific
Carbohydr. Chem., 2012, 38, 303–337 | 329

OH
HO
HO
OH
O
OH
OH O
HO O
HO
O
S
O CdS
HO S
HO O
S
HO HO
O
HO
HO OH
OH
Fig. 23 Mannose-QDs for the detection of E. coli.
interactions between soybean agglutinin and Con A with CdSe/ZnS

glyco-QDs have been monitored using the light scattered at 600 nm.154
Mannose-coated CdS quantum dots (Man-QDs) (Fig. 23) induce lumines-
cent aggregates of E. coli which can be used to detect bacteria in cell
suspensions, containing as few as 104 bacteria per mL. The aggregation
process is due to E. coli cell surface FimH mannose-specific lectins and did
not occur either with an E. coli strain defective in the FimH lectin or with
galactose-coated QDs.155
7.3.2 Live-cell and in vivo animal imaging

Tri-antennary b-galactoside dendrons attached to CdSe/ZnS nanoparticles
were smoothly delivered inside HeLa, kidney, or metastatic lung cancer
cells. The transmembrane efficiency is increased with increasing density of
asialoprotein receptors on cellular surface, indicating a receptor-mediated
endocytosis mechanism.156 Biocompatible chitosan nanoparticles cova-
lently bound to InGaP/ZnS QDs were designed as fluorescent probes for
deep-tissue imaging, improving cellular uptake drastically and allowing
deeper tissue penetration in comparison with PEG–QD nanoparticles.157
Highly biocompatible mannosylated QDs were prepared from commer-
cially available PEG-coated quantum dots in order to interact with man-
nose receptor on macrophages thus tracking macrophage migration and
fate, what is important since macrophages play crucial roles in many
pathophysiological processes.158 QDs bound to biotinylated polymeric LeX
or to HIV-1 envelope glycoprotein gp120 were employed to monitor in
330 | Carbohydr. Chem., 2012, 38, 303–337

living cells antigen-binding, entry and fate of gp120-QDs conjugates once
uptaken by cells.159
Niikura and co-workers studied the uptake of glyco-QDs by digitonin-
permeabilized HeLa cells. It was observed than GlcNAc-QDs accumulate in
the endoplasmic reticulum in the presence of ATP.160 When oligosacchar-
ides161 were attached to the QDs, the fluorescence intensity measured in the
nucleus showed different degree of accumulation depending on the sugar
covering QDs.
Recently, an electrochemiluminescence (ECL) strategy for monitoring
carbohydrate expression on living cells was designed using carbohydrate-
functionalized QDs immobilized on carbon nanotube electrodes. Based on
the specific interactions between lectins and the corresponding carbohy-
drates, this method was used to calculate the average number of mannosyl
residues on living human gastric carcinoma cell surface. This method
has potential application to decode mechanisms of carbohydrate-related
biological processes when dynamic changes in cell surface carbohydrates
are involved.162 Carbohydrate-capped QDs can be exploited for in vivo
targeting as well. Different sugar-PEGylated QDs were tested in vivo for
specific liver sequestration.163 Man- and GalN-QDs accumulated selectively
in the liver due to the presence of specific receptors on hepatic cells.164
8 Conclusions
An impressive development on the synthesis of multivalent glycoconjugates
has taken place in the last few years. The election of the multivalent scaffold
matched with the proper spacer and the right carbohydrate structure plays a
key role in modulating carbohydrate-protein interactions. An overview of
different families of structural assemblies presenting multiple carbohydrate
moieties such as glycodendrimers, glyconanotubes, glyconanoparticles,
show us their multiple biological and medical applications, among which
the design of antitumoral vaccines constitutes a landmark.
Acknowledgements
We thank the Dirección General de Investigación of Spain (CTQ2008
02813) and Junta de Andalucı́a (FQM 134) for financial support. A. O.
thanks University of Seville of Spain for the award of a fellowship.
References
1 R. A. Dwek, Chem. Rev., 1996, 96, 683.
2 B. Ernst and J. L. Magnani, Nat. Rev. Drug Discov., 2009, 8, 661.
3 H. Lis and N. Sharon, Chem. Rev., 1998, 98, 637.
4 M. Mammem, S. K. Choi and G. M. Whitesides, Angew. Chem. Int. Ed., 1998,
37, 2755.
5 Y. C. Lee and R. T. Lee, Acc. Chem. Res., 1995, 28, 321.
6 D. Deniaud, K. Julienne and S. G. Gouin, Org. Biomol. Chem., 2011, 9, 966.
7 R. J. Pieters, Med. Res. Rev., 2007, 27, 796.
8 R. J. Pieters, Org. Biomol. Chem., 2009, 7, 2013.
9 Y-B. Limand and M. Lee, Org. Biomol. Chem., 2007, 5, 401.
Carbohydr. Chem., 2012, 38, 303–337 | 331

10 A. L. Baldini, A. Casnati, F. Sansone and R. Ungaro, Chem. Soc. Rev., 2007,
36, 254.
11 P. Niederhafner, J. Ŝebestı́k and J. Ježek, J. Pept. Sci., 2008, 14, 2.
12 P. Niederhafner, J. Ŝebestı́k and J. Ježek, J. Pept. Sci., 2008, 14, 44.
13 A. Imberty, Y. M. Chabre and R. Roy, Chem. Eur. J., 2008, 14, 7490.
14 Y. M. Chabre and R. Roy, Adv. Carbohydr. Chem. Biochem., 2010, 63, 165.
15 J. J. Lundquist and E. J. Toone, Chem. Rev., 2002, 102, 555.
16 M. N. Alekshun and S. B. Levy, Cell, 2007, 128, 1037.
17 R. J. Pieters, Drug Discov. Today Technol., 2009, 6, e27.
18 A. V. Pukin, H. M. Branderhorst, C. Sisu, C. A. G. M. Weijers, M. Gilbert,
R. M. J. Liskamp, G. M. Visser, H. Zuilhof and R. J. Pieters, ChemBioChem,
2007, 8, 1500.
19 H. M. Branderhorst, R. M. J. Liskamp, G. M. Visser and R. J. Pieters, Chem.
Commun., 2007, 5043.
20 V. Tchesnokova, P. Aprikian, D. Kisiela, S. Gowey, N. Korotkova,
W. Thomas and E. Sokurenko, Infect. Immun., 2011, 79, 3895.
21 S. G. Gouin, A. Wellens, J. Bouckaert and J. Kovensky, ChemMedChem,
2009, 5, 749.
22 S. G. Gouin, J. M. Garcı́a Fernández, E. Vanquelef, F.-Y. Dupradeau,
E. Salomonsson, H. Leffler, M. Ortega-Muñoz, U. J. Nilsson and J. Kovensky,
ChemBioChem, 2010, 11, 1430.
23 C. D. Heidecke and T. K. Lindhorst, Chem. Eur. J., 2007, 13, 9056.
24 M. Hartmann and T. K. Lindhorst, Eur. J. Org. Chem., 2011, 3583.
25 G. Lin, G. Simmons, S. Pohlmann, F. Baribaud, H. Ni, G. J. Leslie,
B. S. Haggarty, P. Bates, D. Weissman, J. A. Hoxie and R. W. Doms, J. Virol.,
2003, 77, 1337.
26 L. Balboa, M. M. Romero, N. Yokobori, P. Schierloh, L. Geffner, J. I. Basile,
R. M. Musella, E. Abbate, S de la Barrera, M. C. Sasiain and M. Alemán,
Immunol. Cell Biol., 2010, 88, 716.
27 S. Sattin, A. Daghetti, M. Theépaut, A. Berzi, M. Saénchez-Navarro,
G. Tabarani, J. Rojo, F. Fieschi, M. Clerici and A. Bernardi, ACS Chem. Biol.,
2010, 5, 301.
28 A. Berzi, J. J. Reina, R. Ottria, I. Sutkeviciute, P. Antonazzo, M. Sánchez-
Navarro, E. Chabrol, M. Biasin, D. Trabattoni, I. Cetin, J. Rojo, F. Fieschi,
A. Bernardi and M. Clerici, AIDS, 2012, 26, 127.
29 P. A. Ledin, F. Friscourt, J. Guo and G.-J. Boons, Chem. Eur. J., 2011, 17, 839.
30 P. Wu, M. Malkoch, J. N. Hunt, R. Vestberg, E. Kaltgrad, M. G. Finn,
V. V. Fokin, K. B. Sharpless and C. J. Hawker, Chem. Commun., 2005, 5775.
31 B. Mokhtari and K. Pourabdollah, J. Incl. Phenom. Macrocycl. Chem., 2012,
73, 1.
32 D. Arosio, M. Fontanella, L. Baldini, L. Mauri, A. Bernardi, A. Casnati,
F. Sansone and R. Ungaro, J. Am. Chem. Soc., 2005, 127, 3660.
33 A. Marra, L. Moni, D. Pazzi, A. Corallini, D. Bridi and A. Dondoni, Org.
Biomol. Chem., 2008, 6, 1396.
34 K. Krenek, M. Kuldova, K. Hulikova, I. Stibor, P. Lhotak, M. Dudic,
J. Budka, H. Pelantova, K. Bezouska, A. Fiserova and V. Kren, Carbohydr.
Res., 2007, 342, 1781.
35 S. André, F. Sansone, H. Kaltner, A. Casnati, J. Kopitz, H.-J. Gabius and
R. Ungaro, ChemBioChem, 2008, 9, 1649.
36 F. Sansone, L. Baldini, A. Casnati and R. Ungaro, Supramol. Chem., 2008,
20, 161.
37 S. Cecioni, R. Lalor, B. Blanchard, J.-P. Praly, A. Imberty, S. E. Matthews
and S. Vidal, Chem. Eur. J., 2009, 15, 13232.
332 | Carbohydr. Chem., 2012, 38, 303–337

38 N. Mesaros, P. Nordmann, P. Plesiat, M. Roussel-Delvallez, J. Van Eldere, Y.
Glupczynski, Y. Van Laethem, F. Jacobs, P. Lebecque, A. Malfroot, P. M.
Tulkens and F. van Bambeke, Clin. Microbiol. Infect., 2007, 13, 560.
39 D. Sicard, S. Cecioni, M. Iazykov, Y. Chevolot, S. E. Matthews, J.-P. Praly, E.
Souteyrand, A. Imberty, S. Vidal and M. Phaner-Goutorbe, Chem. Commun,
2011, 47, 9483.
40 S. Cecioni, J.-P. Praly, S. E. Matthews, M. Wimmerová, A. Imberty and S.
Vidal, Chem. Eur. J., 2012, 18, 6250.
41 E. Galante, C. Geraci, S. Sciuto, V. L. Campo, I. Carvalho, R. Sesti-Costa,
P. M. M. Guedes, J. S. Silva, L. Hill, S. A. Nepogodiev and R. A. Field,
Tetrahedron, 2011, 67, 5902.
42 P. Poignard, E. Ollmann Saphire, P. W. H. I. Parren and D. R. Burton, Annu.
Rev. Immunol., 2001, 19, 253.
43 J. Grove and M. Marsh, J. Cell Biol., 2011, 195, 1071.
44 A. M. Sánchez-Navarro, A. Muñoz, B. M. Illescas, J. Rojo and N. Martı́n,
Chem. Eur. J., 2011, 17, 766.
45 J.-F. Nierengarten, J. Iehl, V. Oerthel, M. Holler, B. M. Illescas, A. Muñoz, N.
Martı́n, J. Rojo, M. Sánchez-Navarro, S. Cecioni, S. Vidal, K. Buffet, M.
Durka and S. P. Vincent, Chem. Commun., 2010, 46, 3860.
46 S. Hakomori, Proc. Natl Acad. Sci. USA, 2002, 99, 10231.
47 M. A. Tarp, A. L. Sørensen, U. Mandel, H. Paulsen, J. Burchell, J. Taylor-
Papadimitriou and H. Clausen, Glycobiology, 2007, 17, 197.
48 Tumor-Associated Antigens, eds. O. Gires and B. Seliger, Wiley-VCH, 2009.
49 M. A. Hollingsworth and B. J. Swanson, Nature Rev., Cancer, 2004, 4, 45.
50 L. Peña Icart, V. Fernández-Santana, R. C. Veloso, T. Carmenate, S. Sirois,
R. Roy and V. Vérez-Bencomo, in Carbohydrate-Based Vaccines, ed. R. Roy,
ACS Symp. Ser., Washington DC, 2008, vol. 989, p. 1.
51 R. M. Wilson, J. D. Warren, O. Ouerfelli and S. J. Danishefsky, in Carbo-
hydrate-based Vaccines, ed. R. Roy, ACS Symp. Ser., Washington DC, 2008,
vol. 989, p. 258.
52 Z. Guo and G. J. Boons (Eds), Carbohydrate-Based Vaccines and Immuno-
therapies, John Wiley & Sons, Hoboken, New Jersey, 2009.
53 Z. Guo and Q. Wang, Curr. Opin. Chem. Biol., 2009, 13, 608.
54 A. Liakatos and H. Kunz, Curr. Opin. Mol. Ther., 2007, 9, 35.
55 U. Westerlind and H. Kunz, Chimia, 2011, 65, 30.
56 R. Roy and T. C. Shiao, Chimia, 2011, 65, 24.
57 L. Morelli, L. Poletti and L. Lay, Eur. J. Org. Chem., 2011, 5723.
58 U. Westerlind and H. Kunz, in Carbohydrate Chemistry, eds. A. P. Rauter,
T. Lindhorst, RSC, Cambridge, UK, 2010, vol. 36, 1.
59 P. Niederhafner, M. Reiniš, J. Šebestı́k and J. Ježek, J. Pept. Sci., 2008, 14,
556.
60 R. D. Atronomo and D. R. Burton, Nat. Rev. Drug Discov., 2010, 9, 308.
61 T. A. Gilewski, G. Ragupathi, M. Dickler, S. Powell, S. Bhuta, K. Panageas,
R. R. Koganty, J. Chin-Eng, C. Hudis, L. Norton, A. N. Houghton and P. O.
Livingston, Clin. Cancer Res., 2007, 13, 2977.
62 T. C. Shiao and R. Roy, New J. Chem., 2012, 36, 324.
63 S. Keil, A. Kaiser, F. Syed and H. Kunz, Synthesis, 2009, 1355.
64 T. Becker, A. Kaiser and H. Kunz, Synthesis, 2009, 1113.
65 A. Kaiser, N. Gaidzik, T. Becker, C. Menge, K. Groh, H. Cai, Y. M. Li,
B. Gerlitzki, E. Schmitt and H. Kunz, Angew. Chem. Int. Ed., 2010, 49, 3688.
66 H. Cai, Z.-H. Huang, L. Shi, Y.-F. Zhao, H. Kunz and Y.-M. Li, Chem. Eur.
J., 2011, 17, 6396.
67 B. L. Wilkinson, C. K. Y. Chun and R. J. Payne, Biopolymers, 2011, 96, 137.
Carbohydr. Chem., 2012, 38, 303–337 | 333

68 B. L. Wilkinson, L. R. Malins, C. K. Y. Chun and R. J. Payne, Chem. Comm.,
2010, 46, 6249.
69 S. Bay, S. Fort, L. Birikaki, C. Ganneau, E. Samain, Y.-M. Coı̈c, F.
Bonhomme, E. Dériaud, C. Leclerc and R. Lo-Man, ChemMedChem, 2009, 4,
582.
70 J. Zhu, Q. Wan, D. Lee, G. Yang, M. K. Spassova, O. Ouerfelli, G.
Ragupathi, P. Damani, P. O. Livingston and S. J. Danishefsky, J. Am Chem
Soc., 2009, 131, 9298.
71 S. J. Danishefsky, G. Ragupathi, P. O. Livingston, J. Zhu, K. Iyer, G. Yang
and P. Sabbatini, 2011, WO 2011/156774 A2.
72 I. Jeon, D. Lee, I. J. Krauss and S. J. Danishefsky, J. Am. Chem. Soc., 2009,
131, 14337.
73 I. J. Krauss, J. G. Joyce, A. C. Finnefrock, H. C. Song, V. Y. Dudkin, X. J.
Geng, D. Warren, M. Chastain, J. W. Shiver and S. J. Danishefsky, J. Am.
Chem. Soc., 2007, 129, 11042.
74 J. G. Joyce, I. J. Krauss, H. C. Song, D. W. Opalka, K. M. Grimm, D. D.
Nahas, M. T. Esser, R. Hrin, M. Feng, V. Y. Dudkin, M. Chastain, J. W.
Shiver and S. J. Danishefsky, Proc. Natl. Acad. Sci. USA, 2008, 105, 15684.
75 S.-K. Wang, P.-H. Liang, R. D. Astronomo, T.-L. Hsu, S.-L. Hsieh, D. R.
Burton and C.-H. Wong, Proc. Natl. Acad. Sci. USA, 2008, 105, 3690.
76 A. Kabanova, R. Adamo, D. Proietti, F. Berti, M. Tontini, R. Rappuoli and
P. Costantino, Glycoconjugate J., 2010, 27, 501.
77 M. C. Galan, D. Benito-Alifonso and G. M. Watt, Org. Biomol. Chem., 2011,
9, 3598.
78 V. Pozsgay, Curr. Top. Med. Chem., 2008, 8, 126.
79 T. Lipinski, P. I. Kitov, A. Szpacenko, E. Paszkiewicz and D. R. Bundle,
Bioconjugate Chem., 2011, 22, 274.
80 H. Wang, L. Gu, Y. Lin, F. Lu, M. J. Meziani, P. G. Luo, W. Wang, L. Cao
and Y.-P. Sun, J. Am. Chem. Soc., 2006, 128, 13364.
81 P. G. Luo, H. Wang, L. Gu, F. Lu, Y. Lin, K. A. Christensen, S.-T. Yang and
Y.-P. Sun, ACS Nano, 2009, 3, 3909.
82 L. Gu, P. G. Luo, H. Wang, M. J. Meziani, Y. Lin, L. M. Veca, L. Cao, F. Lu,
X. Wang, R. A. Quinn, W. Wang, P. Zhang, S. Lacher and Y. P. Sun,
Biomacromolecules, 2008, 9, 2408.
83 M. K. Müller and L. Brunsveld, Angew. Chem., Int. Ed., 2009, 48, 2921.
84 T. Horlacher and P. H. Seeberger, Chem. Soc. Rev., 2008, 37, 1414.
85 S. Park, M.-R. Lee and I. Shin, Chem. Soc. Rev., 2008, 37, 1579.
86 C. Y. Wu, P.-H. Liang and C.-H. Wong, Org. Biomol. Chem., 2009, 7, 2247.
87 H. M. Branderhorst, R. Ruijtenbeek, R. M. J. Liskamp and R. J. Pieters,
ChemBioChem, 2008, 9, 1836.
88 R. M. de Lederkremer and R. Agusti, Adv. Carbohydr. Chem. Biochem., 2009,
62, 311.
89 R. A. Field, P. Andrade, V. L. Campo, I. Carvalho, B. Y. M. Collet, P. R.
Crocker, M. Fais, R. Karamanska, B. Mukhopadhayay, S. A. Nepogodiev, A.
Rashid, M. Rejzek, D. A. Russell, C. L. Schofield and R. M. van Well, in Petite
and Sweet: Glyco-Nanotechnology as a Bridge to New Medicines, eds. X. Huang,
J. J. Barchi, Jr, ACS Symp. Ser., Washington DC, 2011, vol. 1091, p. 143.
90 R. Karamanska, J. Clarke, O. Blixt, J. I. MacRae, J. Q. Zhang, P. R. Crocker,
N. Laurent, A. Wright, S. L. Flitsch, D. A. Russell and R. A. Field, Glycoconj.
J., 2008, 25, 69.
91 M. Fais, R. Karamanska, S. Allman, S. A. Fairhurst, P. Innocenti, A. J.
Fairbanks, T. J. Donohoe, B. G. Davis, D. A. Russell and R. A. Field, Chem.
Sci., 2011, 2, 1952.
334 | Carbohydr. Chem., 2012, 38, 303–337

92 M. Marradi, M. Martı́n-Lomas and S. Penadés, Adv. Carbohydr. Chem. Bio-
chem., 2010, 64, 211.
93 R. Ojeda, J. L. de Paz, A. G. Barrientos, M. Martin-Lomas and S. Penadés,
Carbohydr. Res., 2007, 342, 448.
94 P. Ghosh, G. Han, M. De, C. K. Kim and V. M. Rotello, Adv. Drug Deliv.
Rev., 2008, 60, 1307.
95 S. C. McBain, H. H. P. Yiu and J. Dobson, Int. J. Nanomed., 2008, 3, 169.
96 H. Yang, H. Li and X. Jiang, Microfluid. Nanofluid., 2008, 5, 571.
97 S. Parveen, R. Misra and S. K. Sahoo, Nanomed. Nanotechnol. Biol. Med.,
2012, 8, 147.
98 K. B. Hartman, L. J. Wilson and M. G. Rosenblum, Mol. Diagn. Ther., 2008,
12, 1.
99 E. Boisselier and D. Astruc, Chem. Soc. Rev., 2009, 38, 1759.
100 O. Martı́nez-Ávila, K. Hijazi, M. Marradi, C. Clavel, C. Campion, C. Kelly
and S. Penadés, Chem. Eur. J., 2009, 15, 9874.
101 O. Martı́nez-Ávila, L. M. Bedoya, M. Marradi, C. Clavel, J. Alcamı́ and S.
Penadés, ChemBioChem, 2009, 10, 1806.
102 J. J. Barchi, Jr., in Petite and Sweet: Glyco-Nanotechnology as a Bridge to New
Medicines, eds. X. Huang, J. J. Barchi Jr, ACS Symp. Ser., Washington DC,
2011, vol. 1091, p. 161.
103 I. Garcı́a, J. Gallo, M. Marradi, and S. Penadés, in Engineered Carbohydrate-
Based Materials for Biomedical Applications: Polymers, Surfaces, Dendrimers,
Nanoparticles, and Hydrogels, ed. R. Narain, John Wiley & Sons, 2011, p. 213.
104 Z. Shervani and Y. Yamamoto, Carbohydr. Res., 2011, 346, 651.
105 J. M. de la Fuente, A. G. Barrientos, T. C. Rojas, J. Rojo, J. Cañada,
A. Fernández and S. Penadés, Angew. Chem. Int. Ed., 2001, 40, 2257.
106 J. M. de la Fuente and S. Penadés, Glycoconj. J., 2004, 21, 149.
107 A. Carvalho de Souza, D. N. Ganchev, M. M. E. Snel, J. van der Eerden, J. F.
G. Vliegenthart and J. P. Kamerling, Glycoconj. J., 2009, 26, 457.
108 J. L. Santos, A. C. de Souza, F. J. Canada, S. Martı́n-Santamarı́a, J. P.
Kamerling and J. Jiménez-Barbero, ChemBioChem., 2009, 10, 511.
109 A. J. Reynolds, A. H. Haines and D. A. Russell, Langmuir, 2006, 22, 1156.
110 N. Nagahori, M. Abe and S. I. Nishimura, Biochemistry, 2009, 48, 583.
111 W.-T. Chien, C.-C. Yu, C.-F. Liang, C.-H. Lai, P.-C. Lin, C. C. Lin, in Petite
and Sweet: Glyco-Nanotechnology as a Bridge to New Medicines, eds.
X. Huang, J. J. Barchi, Jr, ACS Symp. Ser., Washington DC, 2011, vol. 1091,
p. 15.
112 X. Wang, L.-H. Liu, O. Ramström and M. Yan, Exp. Biol. Med., 2009, 234,
1128.
113 S. Takae, Y. Akiyama, H. Otsuka, T. Nakamura, Y. Nagasaki and K.
Kataoka, Biomacromol., 2005, 6, 818.
114 C. L. Schofield, R. A. Field and D. A. Russell, Anal. Chem., 2007, 79, 1356.
115 C. L. Schofield, B. Mukhopadhyay, S. M. Hardy, M. B. McDonnell, R. A.
Field and D. A. Russell, Analyst, 2008, 133, 626.
116 M. B. Thygesen, J. Sauer and K. J. Jensen, Chem. Eur. J., 2009, 15, 1649.
117 M. B. Thygesen and K. J. Jensen, in Petite and Sweet: Glyco-Nanotechnology
as a Bridge to New Medicines, eds. X. Huang, J. J. Barchi, Jr, ACS Symp. Ser.,
Washington DC, 2011, vol. 1091, p. 37.
118 M. Reynolds, M. Marradi, A. Imberty, S. Penadés and S. Pérez, Chem. Eur. J,
2012, 18, 4264.
119 A. G. Barrientos, J. M. de la Fuente, M. Jiménez, D. Solı́s, J. Cañada, M.
Martı́n-Lomas and S. Penadés, Carbohydr. Res., 2009, 344, 1474.
120 Z. Ban, C. J. Bosques and R. Sasisekharan, Org. Biomol. Chem., 2008, 6, 4290.
Carbohydr. Chem., 2012, 38, 303–337 | 335

121 X. Jiang, A. Housni, G. Gody, P. Boullanger, M.-T. Charreyre, T. Delair and
R. Narain, Bioconjugate Chem., 2010, 21, 521.
122 M. Chikae, T. Fukuda, K. Kerman, K. Idegami, Y. Miura and E. Tamiya,
Bioelectrochem., 2008, 74, 118.
123 X. Wang, O. Ramström and M. Yan, Anal. Chem., 2010, 82, 9082.
124 X. Wang, E. Matei, L. Deng, O. Ramström, A. M. Gronenborn and M. Yan,
Chem. Commun., 2011, 47, 8620.
125 N. Bogdan, R. Roy and M. Morin, RSC Adv., 2012, 2, 985.
126 S. Nakamura-Tsuruta, Y. Kishimoto, T. Nishimura and Y. Suda, J. Biochem.,
2008, 143, 833.
127 S. Takae, Y. Akiyama, Y. Yamasaki, Y. Nagasaki and K. Kataoka,
Bioconjugate Chem, 2007, 18, 1241.
128 A. Ragusa, I. Garcı́a and S. Penadés, IEEE Trans. Nanobiosci., 2007, 6, 319.
129 P. Eaton, A. Ragusa, C. Clavel, C. T. Rojas, P. Graham, R. V. Durán and
S. Penadés, IEEE Trans. Nanobiosci., 2007, 6, 309.
130 F. Manea, C. Bindoli, S. Fallarini, G. Lombardi, L. Polito, L. Lay, R.
Bonomi, F. Mancin and P. Scrimin, Adv. Mater., 2008, 20, 4348.
131 B. Nolting, J.-J. Yu, G.-Y. Liu, S.-J. Cho, S. Kauzlarich and J. Gervay-Hague,
Langmuir, 2003, 19, 6465.
132 P. M. Enrı́quez-Navas, M. Marradi, D. Padro, J. Angulo and S. Penadés,
Chem. Eur. J., 2011, 17, 1369.
133 M. Marradi, P. Di Gianvincenzo, P. M. Enrı́quez-Navas, O. M. Martı́nez-
Ávila, F. Chiodo, E. Yuste, J. Angulo and S. Penadés, J. Mol. Biol., 2011, 410,
798.
134 X. Li, H. Zhou, L. Yang, G. Du, A. S. Pai-Panandiker, X. Huang and B. Yan,
Biomaterials, 2011, 32, 2540.
135 J. Rojo, V. Dı́az, J. M. de la Fuente, I. Segura, A. G. Barrientos, H. H. Riese,
A. Bernad and S. Penadés, ChemBioChem, 2004, 5, 291.
136 M. M. Kemp, A. Kumar, S. Mousa, E. Dyskin, M. Yalcin, P. Ajayan, R. J.
Linhardt and S. A. Mousa, Nanotech., 2009, 20, 455104.
137 Y.-Y. Chien, M.-D. Jan, A. K. Adak, H.-C. Tzeng, Y.-P. Lin, Y.- J. Chen,
K.-T. Wang, C.-T. Chen, C.-C. Chen and C.-C. Lin, ChemBioChem, 2008, 9,
1100.
138 A. A. Kulkarni, C. Fuller, H. Korman, A. A. Weiss and S. S. Iyer, Bio-
conjugate Chem., 2010, 21, 1486.
139 H. Lee, K. Lee, I. K. Kim and T. G. Park, Biomaterials, 2008, 29, 4709.
140 M. Marradi, D. Alcántara, J. M. de la Fuente, M. L. Garcı́a-Martı́n,
S. Cerdán and S. Penadés, Chem. Commun., 2009, 46, 3922.
141 K. El-Boubbou, D. C. Zhu, C. Vasileiou, B. Borhan, D. Prosperi, W. Li and
X. Huang, J. Am. Chem. Soc., 2010, 132, 4490.
142 K. El-Boubbou, C. Gruden and X. Huang, J. Am. Chem. Soc., 2007, 129,
13392.
143 N. P. Pera, A. Kouki, S. Haataja, H. M. Branderhorst, R. M. J. Liskamp, G.
M. Visser, J. Finne and R. J. Pieters, Org. Biomol. Chem., 2010, 8, 2425.
144 J. M. de la Fuente, D. Alcantara, P. Eaton, P. Crespo, T. C. Rojas,
A. Fernandez, A. Hernando and S. Penades, J. Phys. Chem. B., 2006, 110,
13021.
145 J. M. de la Fuente, D. Alcántara and S. Penadés, IEEE Trans. Nanobio., 2007,
6, 275.
146 Y. Lee, H. Lee, Y. B. Kim, J. Kim, T. Hyeon, H. Park, P. B Messersmith and
T. G. Park, Adv. Mater., 2008, 20, 4154.
147 M. Kamat, K. El-boubbou, D. C. Zhu, T. Lansdell, X. Lu, W. Li and X.
Huang, Bioconjugate Chem., 2010, 21, 2128.
336 | Carbohydr. Chem., 2012, 38, 303–337

148 I. Garciéa, J. Gallo, N. Genicio, D. Padro and S. Penadeés, Bioconjugate
Chem., 2011, 22, 264.
149 J. Gallo, I Garcı́a, N. Genicio, D. Padro and S. Penadés, Biomaterials, 2011,
32, 9818.
150 S. I. van Kasteren, S. J. Campbell, S. Serres, D. C. Anthony, N. R. Sibson and
B. G. Davis, Proc. Natl. Acad. Sci. USA, 2009, 106, 18.
151 F. Y. Cheng, S. P. H. Wang, C. H. Su, T. L. Tsai, P. C. Wu, D. B. Shieh, J. H.
Chen, P. C. H. Hsieh and C. S. Yeh, Biomaterials, 2008, 29, 2104.
152 V. Biju, T. Itoh, A. Anas, A. Sujith and M. Ishikawa, Anal. Bioanal. Chem.,
2008, 391, 2469.
153 C. Earhart, N. R. Jana, N. Erathodiyil and J. Y. Ying, Langmuir, 2008, 24,
6215.
154 P. Babu, S. Sinha and A. Surolia, Bioconjug. Chem., 2007, 18, 146.
155 B. Mukhopadhyay, M. B. Martins, R. Karamanska, D. A. Russell and R. A.
Field, Tetrahedron Lett., 2009, 50, 886.
156 C. T. Chen, Y. S. Munot, S. B. Salunke, Y. C. Wang, R. K. Lin, C. C. Lin,
C. C. Chen and Y. H. Liu, Adv. Funct. Mater., 2008, 18, 527.
157 M. G. Sandros, M. Behrendt, D. Maysinger and M. Tabrizian, Adv. Funct.
Mater., 2007, 17, 3724.
158 Y. Higuchi, M. Oka, S. Kawakami and M. Hashida, J. Control. Release, 2008,
125, 131.
159 A. Cambi, D. S. Lidke, D. J. Arndt-Jovin, C. G. Figdor and T. M. Jovin, Nano
Lett., 2007, 7, 970.
160 K. Niikura, T. Nishio, H. Akita, Y. Matsuo, R. Kamitani, K. Kogure, H.
Harashima and K. Ijiro, ChemBioChem., 2007, 8, 379.
161 K. Niikura, S. Sekiguchi, T. Nishio, T. Masuda, H. Akita, Y. Matsuo, K.
Kogure, H. Harashima and K. Ijiro, ChemBioChem., 2008, 9, 2623.
162 E. Han, L. Ding, S. Jin and H. X. Ju, Biosens. Bioelectron., 2011, 26, 2500.
163 M. Ishikawa and V. Biju, Prog. Mol. Biol. Transl. Sci., 2011, 104, 53.
164 R. Kikkeri, B. Lepenies, A. Adibekian, P. Laurino and P. H. Seeberger, J. Am.
Chem. Soc., 2009, 131, 2110.
Carbohydr. Chem., 2012, 38, 303–337 | 337

Glycotransporters for gene delivery
Carmen Ortiz Mellet,a José M. Garcı́a Fernándezb and
Juan M. Benito*b
DOI: 10.1039/9781849734769-00338
Nucleic acids, from whole genes to short RNA sequences, have emerged as
attractive sources of therapeutic and biotechnological tools. Their highly
specific and predictable mode of action theoretically allow for either sti-
mulating or silencing the expression of any protein. Due to their poor
cellular uptake and rapid degradation in biological media, successful
applications critically depend on the development of efficient purpose-con-
ceived carriers that protect and deliver them into their target cells. Because of
their natural ability to infect cells, modified viruses have been long time
considered as the vehicles of choice. However, given the inherent difficulties
and risks associated to virus manipulation, artificial (non-viral) carriers are
taking over. Within few years, literally hundreds of non-viral nucleic acid
shuttling devices have flourished, based on the most bewildering array of
concepts and materials. Carbohydrates have proven particularly valuable
components for the bottom-up rational design of many of these artificial
nucleic acid carriers. Their structural diversity and biocompatibility, their
unique scaffolding features and targeting capabilities, and their highly flex-
ible chemical functionalization have untapped delirious possibilities where
chemist creativity is the only limit. This review will survey the recent
advances in the design of artificial saccharide-based and saccharide-func-
tionalized gene delivery systems, focussing on the trending topics and stra-
tegies implemented to evade the physiological barriers that hamper nucleic
acid delivery into cells. Emphasis will be laid on the synergic combination of
chemical synthesis and nanotechnology to correlate architectural features of
the glycotransporters with nucleic acid delivering capabilities, feeding back
newly optimized designs. Though major achievements of glycotransporters
in gene delivery are still to come, some of them are already being marketed as
investigational carriers or even successfully clinically trialed.
1 Introduction
It is hardly conceivable for the pharmaceutical industry nowadays, investing
a vast amount of resources on high-throughput drug discovery techno-
logies, that the strategy to bring up new drugs could rely on the design of a
single candidate programmed to develop the required task through a
rational set of rules. And that is precisely what made nucleic acids (genes,
oligonucleotides, aptamers, ribozymes, DNAzymes, or small interfering
RNAs) an attractive source of therapeutic and biotechnological agents.
Their intimate structure-activity relationship and highly specific mode of
a
Departamento de Quı´mica Orgánica, Facultad de Quı´mica, Universidad de Sevilla,
Profesor Garcıá González 1, E-41012 Sevilla, Spain
b
Instituto de Investigaciones Quı´micas, CSIC - Universidad de Sevilla, Ame´rico Vespucio 49,
Isla de la Cartuja, E-41092 Sevilla, Spain. E-mail: juanmab@iiq.csic.es
338 | Carbohydr. Chem., 2012, 38, 338–375

action theoretically permit, when using appropriate designs, to exploit the
cellular machinery in a predictable fashion to either stimulate or silence the
expression of virtually any protein, with reduced potential for toxicity and
fewer side effects. The far-reaching implications on gene therapy and bio-
technology are out of question. The genetic and epigenetic information
obtained from biomedical research could lead to novel research tools,
diagnostic methods and treatments. However, the nucleic acid poor cellular
uptake, due to their large size and anionic charge, and rapid degradation in
biological media turns this conceptually simple strategy into a chimera.
While conventional drugs consist of a formulation of a bioactive species
and a carrier, accounting the former for most of the sophistication of the
design, the success of gene therapy critically depends on the conception of
efficient and safe delivery systems enabling the cargo to reach its target.
Because of their natural ability to infect cells, viruses were the intuitive
choice to deliver genes at the right spot. Several recombinant non-repli-
cating viruses have proved very efficient at promoting genetic material
uptake and expression;1 however, several drawbacks (e.g. immunogenicity,
toxicity and quality of scaled-up production) have tempered the initial
prospects.2 To date, neither the FDA nor the EMA have yet approved any
viral-vector-based gene delivery therapeutic due to these concerns (only a
couple of them are approved by the Chinese State Food and Drug
Administration).3,4 Alternatively, non-viral-based carriers have gathered
momentum. Synthetic materials can be structured using bottom-up
approaches to tailor the desired functional features in terms of bioavail-
ability and biocompatibility, nucleic acid complexation, protecting and
packing capabilities, or membrane-crossing and targeting abilities.
As the understanding of the self-assembling processes involved in
aggregate formation between nucleic acids and synthetic vectors increases,
the range of designed biomaterials for this purpose open wider.5,6 Most of
these non-viral nucleic acid vectors fall within the category of cationic lipids
or polymers (Figs. 1A and 1B, respectively), featuring functional groups
that electrostatically neutralize nucleic acids and cooperatively promote
compaction into colloidal nanoparticles (termed lipo- and polyplexes,
respectively) with increased metabolic stability and membrane permeability.
Nanoparticle stability and transfection efficiency largely depend on the
structural features of the cationic vector, the type of nucleic acid to be
delivered and the relative formulation ratio. Non-viral vectors are, in
principle, less prone to elicit the immune system response. However, they
may trigger inflammatory response and their delivery efficiency and selec-
tivity, despite few exceptions,7 are far from that of their viral counterparts.
This is not surprising considering that most of the first generation non-viral
candidates were not purposely conceived for such task, but just off-the-shelf
materials, such as poly-L-lysine (PLL) or polyethylenenimine (PEI), for
which gene delivery capabilities were incidentally discovered. Their densely-
charged nature does not usually furnish them with the optimal bio-
compatibility and cytotoxicity profiles.8
Recent efforts in the field of non-viral vectors have focused on approaches
aimed at overcoming the physiological barriers to gene delivery.9,10
Hypothesis-driven designs have been implemented aimed at regulating
Carbohydr. Chem., 2012, 38, 338–375 | 339

A NH3 H2
poly-L-lysine (PLL) N NH3
H3N N
n o
branched
O O polythyleneimine
H NH
N NH3 (bPEI)
RHN N m
H
O H3N
n linear polyethyleneimine
(lPEI)
H
N NH2
NH3 NH3 H3N N
n H2
polyamidoamine dendrimer and dendripolymer

(PAMAM)
O
H H
N N N N CORE
H 3N N N
H
O 2 2 O
2
B O
O O
N DOTAP DOPE O O
O
H 2N O P O
O O O
O NH3
O
O NH2
H2
N
O N
H
H 3N NH2
DOSPA
C OH
OH O
NH3 NH3
HO O O HO OH OH O OH
HO O HO O OHO
O HO
H 3N O OH O
OH n OH HO
chitosan HO
OH O
cyclodextrin HO
OH O
schizophyllan OH
HO O OH
HO O O
OH HO O
OH OH OH OHO
O HO O HO OHO
HO O HO
O O
O n = 1–3
OH OH n O
OH HO
n
Fig. 1 Structure of some representative (A) cationic polymers, (B) lipids and (C) carbohy-
drates used for gene delivery. DOTAP: 2,3-dioleoyloxy trimethylammonium propane; DOPE:
2,3-di(oleolyloxy)propyl phosphatidyl ethanolamine; DOSPA: 2,3-dioleoyloxy-N-[2-(spermi-
necarboxamido)ethyl]-N,N-dimethyl-1-propaniminium trifluoroacetate.
340 | Carbohydr. Chem., 2012, 38, 338–375

nucleic acid-carrier condensation,11 enhancing bioavailability and bio-
compatibility of their assemblies,12 shielding aggregates from non-specific
interactions and avoiding off-target delivery,13 promoting selective cellular
uptake (e.g. receptor-mediated)14 or inducing stimuli-responsive cargo
release.15 Carbohydrates have proven particularly useful towards these
goals. Similarly to polyethyleneglycol (PEG) coating (pegylation),16 glyco-
coating can sterically shield the positively charged surface of colloidal
aggregates by improving solvation, thus preventing from non-specific
interactions with intra- and extracellular components. Glycoconjugates can
also be specifically recognized by cell membrane receptors and intracellular
trafficking machineries, whose expression depends not only in cell type but
also in cell developmental state.17,18 Nature already offers a number of
examples that illustrate the potential of carbohydrate polymers as non-viral
nucleic acid vectors, chitosan (Fig. 1C) being probably the most exhaus-
tively investigated.19
Non-viral gene delivery has benefited from the large structural and
functional diversity of carbohydrates, offering broad opportunities to
interfere and manipulate gene transfer capabilities of first generation non-
viral vectors. Moreover, the inherent flexibility of carbohydrate and gly-
coconjugate chemistry has enlarged the range of functional materials that
can be rationally conceived. Systematic studies on structure-function rela-
tionships of such synthetic materials have shed light on the impact that
carrier composition, hydrophobic/hydrophilic balance or charge and
functional group distribution exert on nucleic acid transfer ability. This
review will critically highlight the trending topics on the exploitation of the
structural and functional features of carbohydrates in the bottom-up design
of non-viral gene delivery systems, focussing on three separate aspects: i) the
effect of glycosylation in the performance of first generation non-viral gene
vectors, ii) the de novo design of carbohydrate-based cationic polymers as a
tool to create functional materials with tailored nucleic acid shuttling cap-
abilities, and iii) the design of molecularly well-defined carbohydrate-based
conjugates using pre-organized molecular scaffolds as a mean to intimately
correlate molecular structure with gene transfer capabilities. Due to space
limitations, those approaches based on the use of naturally occurring car-
bohydrate polymers will only be succinctly commented.20
2 Carbohydrate-grafted cationic polymers and lipids

Despite their undisputable investigational utility, manipulation of the
functional features of many of the first generation non-viral vectors is not
an easy task. Cationic polymers, while structurally diverse (e.g. PEI,
PAMAM, PLL, or chitosan; see Fig. 1), are usually very restrained in terms
of precise structural modification. It is known that the polymer molecular
weight and their polydispersity or branching degree significantly influences
their physicochemical properties, delivery efficiency, and toxicity profiles.21
Improved polymerization technologies have rendered low polydisperse
materials,22 but finely tuning these parameters is very rarely sufficient to
achieve clinically useful performances.23 The situation with cationic lipids
does not significantly differ.
Carbohydr. Chem., 2012, 38, 338–375 | 341

Alternatively, the gene transfer capabilities of cationic polymer and lipid
carriers may be manipulated by decoration, either via covalent or supra-
molecular linkages, with additional functional elements. For instance,
pegylation has successfully been exploited to coat transfectious nano-
particles with a hydrophilic shield that enhances stability and biocompat-
ibility by preventing non-specific interactions with biological components.24
Grafting functional elements driving carrier-nucleic acid condensation,25
receptor-mediated nanoparticle cell uptake,26 endosome release27 or cel-
lular14,28 and nuclear29 targeting and trafficking has also been investigated.
Vitamins,30 peptides,31 proteins32,33 or antibodies34 have been used for these
purposes, but carbohydrates probably offer the wider structural diversity
and functional versatility in this regard. This is illustrated by the large array
of glycosylated non-viral carriers reported to date and the diversity of
synthetic strategies to prepare them. From the chronological point of view,
nucleic acid carrier glycosylation was originally conceived to elicit targeted
delivery towards cell membrane specific receptors.35 However, a number of
other potential roles of carbohydrates (e.g. nucleic acid complexation
modulators, nanoparticle biocompatibilizers or active intracellular trans-
port enhancers) have been also investigated to improve the overall perfor-
mance of artificial gene vectors.
A Glycomodulating nucleic acid condensation

Upon designing novel synthetic nucleic acid carriers, condensation and
packing is probably the aspect that receives the lesser attention, despite the
robust evidences of its relevance.36 Electrostatic (non-hierarchical) inter-
action usually leads to suboptimal nucleic acid-carrier packing and con-
siderable nanoparticle dispersity. It has been shown that cationic polymer
glycosylation may serve to modulate this process. Aigner and co-workers37
have recently observed that covalent (oligo)maltose grafting of PEI alter
pDNA and siRNA condensation thermodynamics and nanoparticle stabi-
lity, consequently affecting a number of other features (vide infra, heading
2B). Strand and co-workers have found that the kinetics of DNA com-
plexation and release can be modulated by re-engineering the glycosyl
residues decorating chitosan backbone.38 Sakurai and co-workers succeeded
at accelerating the complex formation kinetics of cholesterol-grafted schi-
zophillan by supramolecularly decorating the cationic polymer with cyclo-
dextrins (Fig. 2).39 Interestingly, glycomodulation of nucleic acid-carrier
interaction is not limited to cationic polymers. Pitard and Lehn have
designed a series of cationic lipids inspired in the structure of RNA-binding
aminoglycosides (Fig. 3), their lipoplexes exhibiting enhanced colloidal
stability as compared with conventional cationic lipids. Remarkably, lipo-
plex particle features and performance could be related to lipid aminogly-
coside structure and siRNA binding capabilities, finely illustrating the
potential of rational design in nucleic acid delivery.40
B Glycomodulating nanoparticle biocompatibility and bioavailability

Structural modification of a gene carrier does not only affect a particular
property, but simultaneously several features. Ascertain the precise effect of
each structural change is, thus, complicated. For instance, oligomaltosylation
342 | Carbohydr. Chem., 2012, 38, 338–375

Fig. 2 Schematic representation of bCD-enhanced cholesterol-appended schizo-
phyllan antisense oligonucleotide (ON) complex formation.39
OH
HO O HO OH O 6
H 3N O
HO
HO N 7
O O H N 6
H 3N NH3
DOSK O 7
OH
HO O
HO
H 3N
H 3N
O
O NH3
O
HO OH
O 6
O OH
N 7
O H N 6
H 3N OH
OH O 7
DOSP
Fig. 3 Structure of some lipidic aminoglycoside derivatives (dioleyl succinyl kanamycin A,

DOSK, and dioleyl succinyl paromomycin, DOSP) and a representative Cryo-TEM micro-
graph of their siRNA complexes (Reproduced from ref. 40a).
of PEI did not only affect nucleic acid binding dynamics, but also physico-
chemical features of the nanocondensates in a carbohydrate density-
dependent fashion.37 (Oligo)maltosylation render complexes that become
largely unaffected by the presence of serum proteins, far less toxic, and exploit
different cell uptake mechanisms as compared to non-maltosylated PEI.
Carbohydr. Chem., 2012, 38, 338–375 | 343

The potential of carbohydrates to modulate the suboptimal biocompat-
ibility of PEI has been also demonstrated by Yu and Wang with a strategi-
cally distinct approach. By crosslinking short PEI chains (PEI600) using
different cyclodextrins (CDs) they built a series of cationic polymers exhi-
biting better DNA-transfecting capabilities than large PEI polymers (e.g.
branched PEI, 25 kDa), while maintaining the low cytotoxic profile of the
shorter polymers (Fig. 4).41 Tang and co-workers have implemented a
number of modifications on this design to achieve targeted delivery of a gene
encoding the epidermal growth factor to folate receptors and enhanced
uptake assisted by cell penetrating peptides (e.g. octaarginine). Preliminary
experiments in vivo demonstrate enhanced antitumor effects on mice as
compared to bPEI (25 kDa) and non-targeted CD-PEI600 polymers.42 Li and
co-workers have exploited the molecular inclusion features of CDs in struc-
turally related polymers to supramolecularly graft PEG chains for improving
their stability under physiological conditions. Such stealth nanoparticles were
less efficient at delivering pDNA but, remarkably, exhibited higher siRNA-
mediated silencing efficiency than commercial formulations.43
Arima and Uekama have conducted parallel studies on PAMAM den-
drimers, which gene delivery efficiency is circumscribed to dendritic gen-
erations for which toxicity is a concern (W G3).44 They found that covalent
grafting of CDs on low generation PAMAM (e.g. G2) resulted in conjugates
maintaining a low toxic profile while boosting transfection capabilities to
rival that of higher generation PAMAM standards,45 a result explained
on the basis of the ability of CDs to interact with membrane components.
Fig. 4 Schematic structure of the folic acid or octaarginine-grafted bCD-PEI600 cationic

polymers reported by Tang and Xu.42b
344 | Carbohydr. Chem., 2012, 38, 338–375

Based in this concept, the authors developed high efficiency carriers for
DNA and RNA delivery.46 Further glycoengineering led Arima and co-
workers to decorate their PAMAM-CD conjugates with saccharidic
antennae. While the initial goal was selective targeting (vide infra, heading
2C), they also found that certain degrees of glycosylation resulted in serum-
resistant polyplexes.47 A similar suppression of serum-induced nanoparticle
aggregation was noted by Sato and co-workers upon glycodecoration of
chitosan with mannosyl and lactosyl residues.48
A number of other polymeric scaffolds have been glycocoated in order to
modulate their biocompatibility and bioavailability with gene delivery
purposes. Carbon nanotubes,49 gold nanoparticles,50 quantum dots,51 or
(pseudo)rotaxanes52 are some relevant examples that further provide
peculiar functionalities to the gene carrier. In addition, glycomodulation of
the pharmacodynamics (e.g. gene entrapment and controlled release) of self-
assembled gene-embedding multilayers has been shown a promising strat-
egy for tissue engineering.53
C Glycomodulating nanoparticle receptor-mediated cell uptake and

intracellular trafficking
While non-viral carrier glycosylation dramatically influence a number of
carrier capabilities, it was certainly carbohydrate specific recognition by
biological receptors that first called the attention. In fact, the first ever
reported evidence of specific non-viral gene delivery used a galactose-ter-
minated PLL carrier targeted to the asialoglycoprotein receptors (ASGPr)
on hepatocyte cell membrane.35 Since then, the topic has received a sus-
tained interest. Galactose (or lactose) and mannose have been mostly
exploited as glycoligands due to their hepatic and immune system cell tar-
geting abilities. But also glucosamine, targeting vimentin,54 and galactosa-
mine55 have been recently used. Glycotargeted versions of the most usual
cationic polymers (including PEI,56 PLL,57 or PAMAM)58 and lipids59 have
been described. The increase in gene expression when using glycoligands is
associated to the improved receptor-mediated uptake of the carriers, but
only to a certain extent. As above outlined, many other features of nucleic
acid-carrier interaction and nanoparticle pharmacodynamics can be altered
upon glycosylation, which makes it difficult to quantify the precise role of
the saccharide antennae. For instance, Behr and co-workers demonstrated
that although the ASGPr is involved in the internalization of galactosylated
PEI by hepatocytes, this is not sufficient to achieve specificity since other
uptake mechanisms simultaneously operate (Scheme 1). Thus, relevant
transfection levels were measured upon addition of ASGPr competing
agents as well as in ASGPr-devoid cells.56a In other cases it was observed
that even though the glycosylated carriers were recognized by the target
receptors, endocytosis did not take place or, if it did, the resulting endo-
somes were unproductive.47a
Glycoligands may also exert a function in controlling the fate of the
internalized nucleic acid-carrier nanocomplexes in a way that remain to be
unveiled. Monsigny and co-workers showed that cytosolic and nuclear
lectins might be involved in intracellular trafficking and nuclear import60
and latter coined the term glycofection to refer to transfection strategies
Carbohydr. Chem., 2012, 38, 338–375 | 345

HO OH OH
O HO OH OH
O
O OH
HO HO
O OH H2
OH OH O N N
HO HO
lactose NaBH3CN OH OH
H2 H NH3
N N
N N H
N NH OH OH
HO
O OH H2
NH3 O N N NH
H HO HO N N
NH N OH OH H2
NH3
lactosylated PEI
H3N N bPEI
ASGPr-mediated
uptake, but not pDNA condensation
exclusive
lactosylated
polyplex
Scheme 1 Synthesis of lactosylated PEI via reductive amination and schematic representation
of the condensation of lactosylated PEI with plasmid DNA into lactosylated polyplexes.56a
based on the glycomodulation of intracellular barriers for gene delivery.61

In a number of cases, lactosylated carriers appeared to induce faster escape
from the late endosome/lysosome than other glycosylated counterparts and
more efficiently mediated nuclear import.62 Though these observations are
still a matter of debate, recent publications by the same group63 and others64
highlight the potential benefits of controlling of intracellular trafficking as a
prerequisite for efficient non-viral gene delivery.
3 De novo designed carbohydrate-based polymers

Efficiency and toxicity of polymeric vectors can be largely (and unexpect-
edly) affected by subtle structural changes, for example in counterion nat-
ure,65 branching degree66 or size,67 which may lead even by batch-to-batch
disparities. This unwanted variability has been overcome to a large extent
by the implementation of novel polymerization techniques, such as con-
trolled polycondensation, radical polymerization and click chemistry reac-
tions, which offer superior control over functional features.68 Through the
judicious choice of monomer structure and polymerization strategies a
number of relevant parameters regarding gene delivery capabilities can be
tailored. In this context, the de novo design of polymeric gene carriers from
carbohydrate-derived building blocks has been very appealing, given the
structural diversity and availability of simple saccharides. This section
surveys a number of examples that illustrate the nucleic acid delivery cap-
abilities of synthetic carbohydrate-based polymers.
A Polyglycoamidoamine-based polymers
One of the most representative examples of the potential of synthetic
glycopolymers at tailoring non-viral gene vectors is the case of
346 | Carbohydr. Chem., 2012, 38, 338–375

polyglycoamidoamines (PGAAs). Polyamidoamines (PAAs) are among the
most accessible entries to structurally controlled polymers with biomedical
applications, their properties being intimately related to structural and
functional features.69 The first insight into the gene delivery capabilities of
carbohydrate-based PAAs (polyglycoamidoamines, PGAAs) were almost
simultaneously reported by Reineke70 and Guan71 (Fig. 5). In both cases the
diversity, availability and biocompatibility of carbohydrates and their
readily conversion to polymerizable building blocks was the main motiva-
tion. Reineke and co-workers focused on imitating the structure of linear
PEI polymers by co-polymerizing their carbohydrate building blocks with a
set of short oligoethyleneimine (OEI) chains. Guan’s design aimed at
mimicking the structure of poly-L-lysine (PLL) by condensing carbohydrate
bricks with lysine segments. Both studies revealed significantly reduced
cytotoxicity of the resulting PGAAs as compared to PEI and PLL,
respectively, while maintained comparable pDNA transfection efficiencies.
Guan and co-workers later oriented their strategy to the elaboration of
protein-resistant biodegradable materials.72 The Reineke’s group on its side
has thoroughly analysed the influence of architectural issues of their PEI-
inspired PGAAs on nucleic acid complexation, nanoparticle topology and
stability, cell uptake, toxicity and gene expression efficiency.73 By screening
a library of cationic PGAAs, (Fig. 6) synthesized from a range of linear
OEIs (different length and number of amino groups) and esterified aldaric
acid and lactone co-monomers (different stereochemistry and number
of hydroxyls), they identified vector candidates that mediated high
gene expression levels in several cell lines (HeLa, H9c2(2-1), HepG2, or
A PGO OPG NPG
PG = protecting group
OPG PGN NPG
HO OH HO OH
OH n OH
O OH OH
B C
OMe O O O
MeO
Cl
OH OH O Cl
MeO2C MeO2C O O O
O O O O
HO HO NH2 NH2
HO OH HO OH
O
H H
N N OMe
H2 N NH2 m = 1 to 4 BocHN N
m H
O O
BocHN
Fig. 5 Schematic representation of the synthesis of polyglycoamidoamines (A) and building

blocks used by Reineke’s70 (B) and Guan’s71 (C) group.
Carbohydr. Chem., 2012, 38, 338–375 | 347

A O
HO OH
O O
HO OH O
N N N N
H H H H
OH OH
B
O OH OH O OH OH
H H H H
N N N N
MeO N n H MeO N H
H H n
OH OH O av. 11-12 OH OH O av. 11-14
D1, n = 1 G1, n = 1
D2, n = 2 G2, n = 2
glucarate-based PGAAs galactarate-based PGAAs
D3, n = 3 G3, n = 3
D4, n = 4 G4, n = 4
O OH OH O OH
H H H H
N N N N
MeO N H MeO N H
H n n
H
OH OH O av. 11-14 OH O av. 11-12
M1, n = 1 T1, n = 1
M2, n = 2 T2, n = 2
mannarate-based PGAAs tartrate-based PGAAs
M3, n = 3 T3, n = 3
M4, n = 4 T4, n = 4
OH O
H H
MeO N N
N N H
H H
O OH av. 12-14
DS, = glucarate
spermine-based PGAAs
TS, = tartrate
Fig. 6 General structure of PGAAs studied by Reineke and co-workers (A) with indication of
their structural diversity (B).
BHK-21). pDNA transfer efficiency peaked for the polymers containing the
longer OIE chains tested (four secondary amino groups).74 Transfection
efficiency favourably compared to that achieved with PEI, but was
accompanied by a far less toxic profile. They hypothesized that the iterative
interruption of OEI charge density by carbohydrate blocks was at the origin
of the reduced toxicity as compared to PEI, though their rapid hydrolysis in
biological media might also account for this merit.75
PGAAs with longer OIE segments did not significantly influenced cell
uptake and overall transgene expression, but significantly improved poly-
plex stability in serum-containing media. Unfortunately, increased toxicity
was also observed.76 Polymers with longer inter-amine spacers (e.g. sper-
mine) also exhibited increased toxicity, highlighting a critical relationship
between charge density distribution and biocompatibility.77 The nature of
the saccharide segment, though modestly, also influenced polymer perfor-
mance, apparently by altering the buffering capabilities of the system.
Since the response of polyplexes at the increase of pH in the endosomes
closely relates to the pDNA release dynamics, the results might indicate
certain folding differences upon pDNA:PGAA complex formation. This
study let identify PGAA G4, combining pentaethylenehexaamine and
galactarate segments (Fig. 6), as the best performing vector candidate,
promoting high transgene expression levels in a range of primary cell lines
(similar to PEI) without causing relevant toxicity. This novel formulation
has been recently licenced to Techulon and is now commercialized as
Glycofectt.78
348 | Carbohydr. Chem., 2012, 38, 338–375

B Amidine- and triazole-linked carbohydrate-based polymers
As an alternative to PGAAs, carbohydrate-based cationic polymers with
amidine79 and, more recently, triazol linking tethers80 have also been
explored. The amidine functionality is protonated at physiological pH,
which might be favourable regarding nucleic acid complexation and stabi-
lization. However polyamidines usually exhibit severe cytotoxicity. Reineke
and Davis observed that this drawback could be overcome by intercalating
saccharidic moieties between the cationic elements while maintaining the
gene transfer capabilities.79 The closer the amidine and the carbohydrate
units are located in the polymer chain, the more efficiently the cytotoxicity is
reduced. Such favourable effect was first observed for trehalose-amidine
copolymers and further confirmed for b-cyclodextrin as the sugar building
block (vide infra, heading 3C).
Trehalose itself has been found to impart some favourable properties
to polymeric gene vectors, such as shielding aggregation of the
resulting polyplexes with proteins and lipids.81 To fully benefit from this,
Reineke and co-workers have devised a strategy to build a new family of
cationic trehalose-embedding polymers that exploits the Cu(I)-catalysed
azide-alkyne cycloaddition (CuAAC) reaction.82 By coupling diazido-
functionalized trehalose units with a diverse set of a,o-dipropargylated
OEIs (Scheme 2), a series of polymers were obtained that allowed the sys-
tematic investigation of architectural effects (OEI size and polymer length)
on pDNA complexation, nanoparticle properties, cellular uptake and
transgene expression. Transfection efficiencies that paralleled that of jetPEI-
based polyplexes were achieved using trehalose ‘‘click’’ polymers with 3-4
ethylenediamine segments.80a Dynamic light scattering revealed that poly-
plexes formulated with larger polymers exhibited the best tolerance
to serum-containing media, but displayed higher toxicity.80b This unfa-
vourable feature was alleviated by the fact that ‘‘click’’ polymers required
far lower N/P ratios, as compared to PGAAs, to achieve similar gene
expression levels.
HO AcO
OH O N3 O
O O O O
OH OH OAc OAc
HO OH OH AcO OAc OAc

HO N3 oligoethyleneimines
trehalose +
n = 1–3
H H
N N H2 N NH2
N N
Boc n H n
O O
i) CuSO4/ascorbate
ii) NaOMe, MeOH
iii) TFA-CH2Cl2
N
O O
N HO H
N O N
N N
O O H n H
OH OH
N N n = 1–3
N
HO OH OH av 50–60
Scheme 2 Synthesis of trehalose-based cationic ‘‘click’’ polymers.80
Carbohydr. Chem., 2012, 38, 338–375 | 349

C Cyclodextrin-based polymers
The application of CD-containing polymers to gene delivery was pioneered
by Davis and co-workers. At the light of the increasing interest on the
biomedical application of CD polymers,83 they conceived a new class of
cationic polymers specifically designed to deliver macromolecular ther-
apeutics.84 Their synthetic strategy was based on the controlled poly-
condensation of bifunctional CD monomers and cationic co-monomers, in
order to yield linear chains with alternating CD and cationic units (CDP in
Scheme 3).85 Electrostatic-driven complexation of the resulting cationic CD
polymers (CDPs) and negatively charged pDNA (B5 kpb) rendered
nanometric polyplexes (polyCDplexes; 100–150 nm) featuring in vitro cell
transfection efficiency comparable to that obtained with polyethyleneimine
(PEI) and Lipofectaminet while preserving a reduced toxicity.
This milestone contribution was followed by a series of reports in which
the effect of structural modifications of CDPs on the gene delivery cap-
abilities were investigated.86–91 The influence of factors such as the CD size
(b- or gCD, in comparison with linear saccharides)87,89 and the distribution
and nature of cationic elements88,90 (their linkages, distances,86 and relative
dispositions),87 as well as the polymer size and polydispersity,89 have been
examined. Structure-activity relationship (SAR) studies concluded that low
molecular weight polymers (ca. 10 kDa, degree of polymerization, DP, 5-8)
with the CD units sufficiently spaced from amidine cationic centres were the
optimal architectures in terms of both high delivery efficiency and low
cytotoxicity (Scheme 3).91
More recently, the same authors have reported a significant improvement
in delivery efficiency by modifying the polymer endings with imidazole
groups.92 Though the precise mechanism leading to this efficiency
enhancement is unclear, the authors argued that imidazole moieties should
impart buffering capacity to CDP-gene polyCDplexes that would prove
Scheme 3 Synthesis of bCD-based cationic polymers (top) with indication of their structural
diversity (bottom).
350 | Carbohydr. Chem., 2012, 38, 338–375

instrumental for efficient endosomal release by virtue of the ‘‘proton
sponge’’ effect.93
Very recently Srinivasachari and Reineke have investigated a versatile
approach towards linear CD-containing cationic polymers for pDNA
delivery by using Cu(I)-catalysed azide-alkyne 1,3-dipolar cycloaddition of
a diazido bCD derivative and a,o-dipropargylated oligoethyleneimines
(Scheme 4). The gene expression profiles of the resulting ‘‘click’’ polymers in
HeLa (human epithelial cervical carcinoma) cells were mostly dependent on
the oligoethylenemine/CD ratio, with the polymers having longer oli-
goethylenemine segments being the better performers.94
The nanoparticles obtained from CDPs upon oligonucleotide com-
plexation can be supramolecularly modified at their surface. For instance,
the inclusion of the adamantane moiety of adamantane-polyethyleneglycol
conjugates (Ad-PEG) into CD cavities of CDP-pDNA complexes creates a
steric shield around the particles that prevents aggregation and non-specific
interactions with biological components.95 Furthermore, polyCDplexes
coated with galactosylated Ad-PEG were shown to exhibit selectivity
towards hepatocytes.96 PolyCDplexes covered with transferrin-modified
Ad-PEG (Ad-PEG-Tf)97,98 transfected luciferase-encoding gene to K562
human myelogeneous leukemia cells with better efficiency than non-targeted
particles (Fig. 7).97 These Tf-targeted gene vectors have shown in vivo
transfection efficiency and selectivity towards different tumor models
(murine99–101 and primates).102 In June 2008, a targeted therapeutic based in
this concept, CALAA-01 (Calando Pharmaceuticals) entered phase I clin-
ical trials. The active ingredient of CALAA-01 is a small interfering RNA
(siRNA). This siRNA inhibits tumour growth via RNA interference to
reduce expression of the M2 subunit of ribonucleotide reductase (R2). The
CALAA-01 siRNA is protected from nuclease degradation within a stabi-
lized polyCDplex termed RONDELt (RNAi-oligonucleotide nanoparticle
delivery). RONDEL is a three part delivery system: the cationic linear CDP,
the AD-PEG surface modification element to increase stability and serum
half-life and the Ad-PEG-Tf targeting ligands that assure targeting of the
nanoparticles to the tissues of interest. The study is directed to adults with
solid tumours refractary to standard-of-cure therapies and is currently
HO OHOH AcOOAc OAc

I I N3 N3
HO OH AcO OAc O O
Boc
i) NaN3, DMF, 80 ºC N
N N
H 1-4 H
ii) Ac2O, Py
i) CuSO4, ascorbate
ii) NaOMe/MeOH
iii) TFA-DCM
O O
N N H2
HO OHOH N
N N N N
H 1-4 H
HO OH N N
av. 40-200
Scheme 4 Synthesis of bCD-based cationic ‘‘click’’ polymers.94
Carbohydr. Chem., 2012, 38, 338–375 | 351

Fig. 7 Schematic representation of the elaboration of the transferrin-targeted nucleic acid-
CDP nanoparticles (RONDELt) developed by Davis and co-workers.77
recruiting participants.103 Alternative targeting ligands (e.g. cancer cell

markers and antibodies) are being investigated.7,104,105
D Diblock glycopolymers
Structural control of polymeric species has experienced a dramatic
improvement with the advent of modern polymerization technologies. For
instance, living radical polymerization enables the synthesis of polymers, co-
polymers or multi-block polymers with highly defined architectures and
broad functional group tolerance. Such polymerization techniques have
already shown promising perspectives in the elaboration of functional gly-
copolymers.106 Ahmed and Narain have recently used radical addition-
fragmentation chain transfer (RAFT) polymerization to synthesize a library
of carbohydrate-pendant cationic polymers, where carbohydrate and
cationic units can be either statistically distributed (random polymers) or
segregated into separate blocks (diblock polymers) (Scheme 5). Precise
control on chain length, monomer ratio and polydispersity allowed the
authors to rationalize the large influence that polymer architecture exerts on
gene delivery capabilities. Statistical copolymers with high molecular weight
produced superior gene expression with lower toxicity as compared to the
corresponding diblock copolymers, both in the presence or absence of
serum.107 On the other hand, diblock copolymers resulted in nanoparticles
352 | Carbohydr. Chem., 2012, 38, 338–375

HO OH HO OH
A
v OH x y OH z
statistic copolymerization
HO OH HO OH
homopolymerization homopolymerization
OH n OH m
diblock polymerization
HO OH
OH OH
OH n m
B O
HO
HO OH
OH
H H OH HN
HO OH
N N H H3N
HO HO OH H3N
N
HO HO
O GAPMA O NH
HO NH
O HN O O HN O O
H
N statistic copolymer
H3N
AEMA O v x y z
CTP/ACVA OH
NH3
HO OH H H3N
N
HO
O NH HO NH
GAPMA O HN O O
HO2C S Ph
CTP/ACVA HO2C S Ph
CN S CN NC n m S
ACVA HO2C N
N CO2H
diblock copolymer
CN
HO2C S Ph
CTP
CN S
Scheme 5 Schematic synthesis of statistical vs diblock copolymers (A) illustrated with the
cationic glycopolymers synthesized by Ahmed and Narain (B).107
that were less prone to unspecifically interact with serum proteins and cell
membrane components. Such shielding features are comparable to those
imparted by conventional PEG grafting on cationic polymers, but probably
more easily modulated.
RAFT polymerization has been also exploited by Reineke and co-
workers for the synthesis of a series of diblock cationic glycopolymers
having the sugar units in the chain (Scheme 6).108 Interestingly, regardless of
the cation-to-carbohydrate balance, diblock polymer-pDNA colloids were
found to be stable in the presence of salt and serum over several hours,
whereas other formulations (PEI or Glycofectt) tend to aggregate. Cell
uptake of the complexes was found to be as efficient as that observed for
PEI or Glycofectt. Although pDNA transgene expression was only mar-
ginal, siRNA-induced gene knockdown efficacy was comparable to that
achieved with Lipofectaminet 2000.
Carbohydr. Chem., 2012, 38, 338–375 | 353

HO
HO O
HO OH
HO O HO
HO HO NH CN
O CPP/ACVA HO O HO OH
HO HO NH O
HO OH HO O HO OH
NH HO OH NH O CO2H
O NH O
O
S
S av. 43
C3H7 S
AEMA
CPP/ACVA
S S CO2H HO
CPP O
HO HO
S CN O HO OH
HO HO
O HO OH NH CN
HO HO
H3N O HO OH NH O
HO
H3N HO OH NH O CO2H
H3N NH O
NH O
NH O
NH O av. 43
O 21–48
S
S
C3H7 S
Scheme 6 Synthesis of diblock cationic glycopolymers by RAFT polymerization described by

Reineke and co-workers.108
The current examples of diblock glycopolymers cannot compete in terms

of gene transfer efficiencies with conventional natural or synthetic glyco-
polymers. Nevertheless, the potential of controlled polymer manipulation
techniques open new opportunities with promising perspectives in the field
of non-viral gene delivery. Targeted delivery for instance, where non-
specific interactions have to be shielded at the same time that specific uptake
is promoted, might be one of those channels.
4 Preorganized glycomaterials for gene delivery

A Non-carbohydrate macrocycle-scaffolded glycoclusters
Appending or inserting carbohydrate moieties into linear and branched
polymeric or oligomeric amines has proven a very versatile strategy to
access gene delivery systems with improved biocompatibility, furtivity,
membrane crossing, and/or targeting abilities.7,73 The intrinsic poly-
dispersity of these glycomaterials and their random conformational prop-
erties make it difficult to undertake a systematic investigation of the
influence of structural modifications on the transfecting properties, how-
ever. Moreover, their generally flexible character may give rise to self-
folding of the polyamine chains,109 which decreases the binding ability
towards DNA and forces the use of higher N/P ratios to achieve full
complexation and protection. Preorganization of the cationic functional
elements onto macrocyclic platforms has the potential to allow controlling
their spatial orientation and, ultimately, the self-assembling behaviour of
discrete architectures to produce nanometric objects that can be pro-
grammed to complex, compact, deliver and release plasmid DNA in a target
cell. Most interestingly, homogeneity can be preserved at the molecular level
354 | Carbohydr. Chem., 2012, 38, 338–375

in structurally related series of compounds by implementing selective che-
mical functionalization methodologies, offering unprecedented opportu-
nities for structure-activity relationship studies.110
Elaboration of macrocyclic-based glycomaterials with gene delivery
capabilities can be conceived through two general tactics, namely glycode-
coration of non-carbohydrate platforms or incorporation of functional
elements for oligonucleotide complexation onto cyclooligosaccharide
backbones. Several non-carbohydrate macrocyclic scaffolds have been
exploited to implement the concept of preorganized gene delivery systems,
including macrocyclic polyamines,111 calixarenes112 and calix[4]resorcinar-
enes (resorcarenes).113 All of them have been largely used for the con-
struction of glycoconjugates that proved extremely useful to assess the
influence of architectural parameters in carbohydrate-protein recognition
processes,114 but only in the last case the formation of transfectious gly-
cocomplexes with DNA after installation of the carbohydrate appendages
has been demonstrated and thoroughly investigated.
The use of glycoresorcarenes in gene delivery was pioneered by Aoyama,
in Japan, in a work originally intended at developing unimolecular mimics
of the cell-membrane sphingoglycolipid clusters.115 For such purpose, the
bowl-shaped calix[4]resorcarene framework was chosen.116 Coupling of
octaamine 1 with the disaccharide (cellobiose, maltose and lactose) lactone
derivatives 2–4 afforded octavalent glycolipid bundles 5–7 bearing four
undecyl tails, with the hydrophilic sugar moieties and the hydrophobic
chains located onto opposite rims of the macrocyle in a well-defined geo-
metry (Scheme 7).
H2N NH2 H2N

NH2
H2N NH2
OH
OH
NH2 O OO O O HO O
O NH2 O
O O + HO O
OH HO O
OH
MeOH 2, Cello
7 44–76% 3, Malto
7 7 4, Lacto
7 1
OR
HO OR
RO OR OH
HO HO OH OH HO
OH OH
OH HO
RO HO O HO OH HO OR
OH
OH NH HO HN
HO NH HN O O HO OH
OH HO H
N N
HO O OO O O
O O OH
OH O O O
RO O OH
OH H
HO N NH OR
HO O HO
HO
5,R= β-D-Glc(Cel8)
6,R= α-D-Glc(Mal8)
7,R= β-D-Gal(Lac8)
Scheme 7 Synthesis of glycoresorcarenes.
Carbohydr. Chem., 2012, 38, 338–375 | 355

(a) (b)
pDNA
glyconanoparticle glycovirus
4–6 nm 50–60 nm
Scheme 8 Schematic representation of the process leading from glycoresorcarenes to glyco-

micelles and to glycoviruses.
Given their amphiphilic character, glycoclusters 5–7 were expected to

form micelle-like aggregates in water. Notwithstanding, they were found to
be not surface-active. A combined GPC and TEM study revealed that they
indeed form micelle-like spherical aggregates, termed glycocluster nano-
particles (GNPs), with a diameter of 4-5 nm and an aggregation number of
B6 (Scheme 8, step a). In marked contrast to usual cases, micellization is
practically irreversible for those systems. Strong intermolecular hydro-
phobic entangling of the alkyl groups, possibly coupled with lateral or
intercluster hydrogen bonding in the glycocluster domain, may be at the
origin of the unusual stability.117 Interestingly, the GNPs were serendipi-
tously found to undergo agglutination with phosphate ions as a glue, which
involves the establishment of polar sugar-to-phosphate interactions, possi-
bly reinforced by inter GNP sugar-sugar interactions. This unexpected
phenomenon led Aoyama’s group to take up DNA, a natural phosphate
polymer, as an interaction partner of GNPs.
The three resorcarene-based glycoclusters 5–7 were able to complex
plasmid DNA to afford transfectious nanometric complexes called glycov-
iruses by the authors (Scheme 8, step b). The results evidenced that at a host/
phosphate (or base) ratio of B0.6 saturation is reached and the size of the
resulting glycovirus remains nearly constant thereof (B54 nm). The surface
(z) potential of the complex, which is negative in the presaturation region,
becomes then neutral. The 0.6:1 (i.e., 12:20) glycocluster:nucleobase ratio
indicates that two particles of hexameric GNPs accommodate in every
helical pitch of the DNA, composed of 10 bp. The glycovirus particles were
capable of delivery a luciferase-encoding reporter gene into cultured cells,
which is remarkable considering the neutral character of the vector. In
general, a strong dependence of the transfecting efficiency on the nature of
the carbohydrate moieties was observed, however, which inversely corre-
lated with the tendency of the corresponding glycoviruses to self-aggregate
(Mal W Lac WW Cel).
The hierarchical process leading from glycoclusters to glyconanoparticles
to glycoviruses warrants that at any molecular or aggregated stage there will
be sugar residues exposed to the external medium, therefore potentially
available to act as ligands of complementary receptors. Actually, the formal
valency must increase with the size of the supramolecular edifice, favouring
recognition by lectins. Thus, in the case of the hepatic cell line HepG2, Lac
glycoviruses exhibited a much higher activity than expected due to the
occurrence of specific interactions between the galactose residues and the
356 | Carbohydr. Chem., 2012, 38, 338–375

asialoglycoprotein receptors present at the cell surface, which could be
exploited for targeted gene delivery. Optimization of the cluster structure
was achieved by decreasing the number of lactose residues onto the
octaamine scaffold and keeping cationic amino groups at the unsubstituted
positions, thereby minimizing the detrimental impact of self-aggregation.
Heterogeneous compounds with an overall (Lac)3(NH3 þ )5 resorcarene
structure exhibited the best properties in term of efficiency of gene delivery
to hepatocytes and cell viability.
B Cyclodextrin-scaffolded gene delivery systems

Modulating the molecular topology and the finite macromolecular asso-
ciation of preorganized systems by acting not only on the head and tail
groups nature and density, but also on the size and shape of the macrocyclic
nucleus offers further opportunities to optimize both transfection efficiency
and cell viability parameters. The use of cyclodextrins (CDs) as macrocyclic
platforms to build preorganized gene delivery systems deserves special
consideration in this context. Cyclodextrins are biocompatible, commer-
cially available nano-objects with two well-differentiated and chemically
distinguishable faces, one limited by the primary hydroxyls and the other by
the secondary ones. Installation of a broad range of functional elements in a
controlled, space-oriented manner can then be achieved through position-
selective and face-selective functionalization methodologies. This chemical
flexibility has been exploited to elaborate macromolecular constructs that
can be programmed to recognize biomacromolecular partners, including
lectins,118 toxins119 and DNA.120
In 2004 O’Driscoll and Darcy reported the first examples of mono-
disperse cyclodextrin-centred polycationic clusters intended for DNA
complexation,121 namely bCD derivatives bearing alkyl and arylamine
antennae on their primary rim and unmodified OH-2 and OH-3 hydroxyls
(e.g. 8), and showed that they were able to promote transfection in vitro to
some extent. This work has been followed by others, including Yannako-
poulou’s,122 Reineke’s123 and Midoux’s groups,124 who proposed the
installation of guanidinium (9)/guanidinioalkylamino (e.g. 10), oligoethy-
leneamine (e.g. 11) or histidine (12)/lysine substituents (13) at the primary
(C-6) positions of CD platforms (Fig. 8). The highly hydrophilic character
of these materials someway hampers the self-organization of the monomers
at the surface of the oligonucleotide chain, however. Consequently, rela-
tively big excesses are necessary to achieve full protection from the envir-
onment. An exception is the histidine-aCD derivative 12, for which it was
hypothesized that the corresponding CD-DNA complexes (CDplexes) are
stabilized through intermolecular inclusion involving the imidazole ring
and the CD cavity. In any case, all these series the CD-based polycations
were able to induce condensation of DNA. In some cases, the resulting
CDplexes exhibited remarkable cell-penetrating and transfection cap-
abilities while featuring much less toxic profiles than commercial cationic
polymers.
The above examples illustrate the suitability of CDs as molecular plat-
forms for the incorporation of cationic elements with a precise spatial
orientation and the potential of the resulting monodisperse polycations as
Carbohydr. Chem., 2012, 38, 338–375 | 357

H
N
NH2
H2N NH2
N NH2
H
NH NH NH
O O O
HO HO HO
OH O OH O OH O
7 7 7
8 9 10
NH3
NH3 NH NH3
NH2
HN O O
N NH3
N H
N NH NH
N O
O O O
HO HO HO
OH OH OH
O O O
7 6 7
11 12 13
Fig. 8 Examples of non-amphiphilic polycationic cyclodextrins assayed as gene vectors.
gene vectors. Most interestingly, the rim anisotropy of the basket-shaped

CD structure further allows accessing compounds with segregated cationic
and lipophilic domains for which, according to the ‘‘facial amphiphilicity’’
concept, biomimetic self-assembly and gene delivery properties could be
expected. Darcy and co-workers first tested this hypothesis by preparing
heterogeneous bCD derivatives bearing thioalkyl chains at the primary rim
and amine-terminated oligoethylene branches of variable lengths at the
secondary hydroxyl positions. The resulting polycationic amphiphilic CDs
(paCDs) were shown to entrap pDNA by forming CDplexes that trans-
fected COS-7 and human hepatocellular liver carcinoma HepG2 cells with
efficiencies and toxicity profiles comparable to Lipofectaminet 2000-based
lipoplexes.125
The possibility of implementing synthetic methodologies allowing full
control of the chemical composition and of the conformational behavior of
the vector let envisage the goal of investigating the influence of structural
modifications at the molecular level on the self-assembling, DNA com-
plexation, cell internalization and transfection properties. With this idea in
mind, a multicollaborative project was settled between several European
laboratories, including those of Defaye (Grenoble, France), Vierling and Di
Giorgio (Nice, France), Ortiz Mellet and Garcı́a Fernández (Sevilla, Spain),
to develop a new family of monodisperse polycationic amphiphilic CD-
based glycomaterials as gene delivery systems. The first candidate was a
bCD derivative with seven cysteaminyl groups at the primary face and
fourteen hexanoyl chains at the secondary rim, accessible in only three steps
in multigram scale after a single column chromatography purification.126
The synthetic sequence starts with the displacement of the bromo groups in
the heptabromo derivative 14 with N-Boc-protected cysteamine (-15). The
fatty acyl chains were put in place by subsequent reaction with hexanoic
anhydride in DMF using DMAP as base promoter (-16). These acylation
conditions warrant full homogeneity of the final compound, avoiding
358 | Carbohydr. Chem., 2012, 38, 338–375

S
S O NHBoc
Br NHBoc
HS O
O O O O
NHBoc 4 2
HO HO O O
HO Cs2CO3, DMF HO DMAP, DMF O O
O O
95% 72%
7
7 7
14 15 16
CIH·H2N i) TFA-H2O
ii) HCl
S NH2·HCl
99%
O O OO
O S
O O OO OO
CIH·H2N S O
O
O O OO
OO O O
O O
O NH2·HCl
S
S O O O
O O
O O O
CIH·H2N OO OO O
O
O
O O S
S O
NH2·HCl
NH2·HCl
17
Scheme 9 Synthesis of the cysteaminyl polycationic amphiphilic cyclodextrin (paCD) 17.
under- and overacylation side products. Final acid hydrolysis of the seven
carbamate groups afforded the target paCD derivative 17 in excellent yield
(Scheme 9).
Electrophoretic mobility experiments on agarose gel indicated that the
amphiphilic cysteamine derivative formed a stable complex with pDNA at
N/P 5 that fully protected the plasmid from degradation by nucleases or
from binding to the intercalating agent etidium bromide. The plasmid was
recovered intact after dissociation of the complex by treatment with sodium
dodecylsulfate (SDS). Further transfection experiments in COS-7 cells
indicated that it was as efficient as branched PEI in promoting transfection
at N/P 5. However, the efficiency was 100-fold lower as compared to bPEI
at N/P 10, the optimal charge ratio for the reference cationic polymer,
although with a much more favorable toxicity profile.
To improve the transfection efficiency, the authors got inspiration from
Nature’s mechanisms of phosphate anion recognition,127 which imply the
concerted action of hydrogen bonding and electrostatic interactions. Thus,
a new vector prototype incorporating a belt of thiourea segments between
the cyclooligosacharide core and the polycationic cluster was designed. The
synthesis implies the nucleophilic addition of the heptacysteaminyl deriva-
tive 17 to Boc-protected 2-aminoethyl isothiocyanate and further acidic
cleavage of the protecting groups, affording the amphiphilic poly-
aminothiourea adduct 18 in 85% overall yield (Scheme 10). The homo-
geneity of the compound was confirmed by combustion analysis, MS and
NMR spectroscopy, which fully demonstrated that the C7 molecular sym-
metry was preserved. The biological results indicated a significant increase
in the transfection efficiency for polyaminothiourea 18 as compared with
polyamine 17, paralleling that for polyplexes formulated with PEI at its
optimal N/P 10 ratio.126 Moreover, transfection efficiency was not
Carbohydr. Chem., 2012, 38, 338–375 | 359

H
ClH·H2N N ClH·H2N
S
HN
NH
S HN S
S O O O
O
O S
HN O O OO
N S OO O
S H O
NH2·HCl ClH·H2N O OO NH2·HCl
O
O OO O O
O i) SCN(CH2)2NHBoc O O H
O O N NH
O S
O O ii) TFA-CH2Cl2 iii) HCl
S O O O S
O O
7 85% S O O
O O O
NH O OO O
17 HN O
O O
O S
S
S NH NH2·HCl
NH2·HCl
S N
N N NH S
S H
H H NH3
HN NH2·HCl
RO O DNA 18
P
R'O O
Scheme 10 Synthesis of the aminoethylthioureido paCD 18.
affected in serum-containing media, an important requisite for in vivo

applications.
The above synthetic strategy had been purposely conceived to be mole-
cular diversity-oriented. In a first series of compounds, the nature of the
spacer between the thiourea and the amino groups was systematically
modified.128 Aromatic spacers proved to be detrimental for transfection
and, moreover, led to a substantial increase in cytotoxicity. In contrast,
aminoalkyl thioureido derivatives were excellent transfection agents, with
the two-carbon spacer being the optimal structure. Their comparative
evaluation indicated that a dendritic presentation of amino groups
separated by ethylene segments provided the best transfection capabilities,
with transfection efficiencies 1000- or 10-fold higher as compared with the
reference polycationic polymer PEI at N/P 5 or 10, respectively (Fig. 9).
Dynamic light scattering (DLS) measurements and transmission electron
microscopy (TEM) images of the CDplexes obtained from amphiphilic
polyaminothiourea-CDs revealed a narrow size distribution. The nano-
complexes are rather regular and have about the optimal size (40–50 nm) for
systemic transfection. Considering that the plasmid alone has a size of
above 1 mm the compaction ratio was about 20-fold, which according to
theoretical calculations implies that they are monomolecular in DNA. A
snail-like ultrastructure can be distinguished in which the electron-dense
dark regions correspond to the DNA chain and the lighter regions to the
self-assembled paCDs. Probably the later are arranged in bilayers, with the
hydrophobic domains facing each other and the cationic clusters pointing to
the oligonucleotide chain, thereby favoring desolvation and condensation of
the plasmid (Fig. 10).
The fact that the transfection efficiencies improved with an increase in the
number of cationic groups over the initial 7 primary centers in bCD deri-
vatives suggested that reversing the facial polarity, by anchoring the
360 | Carbohydr. Chem., 2012, 38, 338–375

log (mg Luc/mg protein)
8
4
S
S
N X PEI NH3 NH3 N
O H N N 5 H
H H NH3
O X=
NH 3 N NH3
O OO N H
H 3
O
log (mg Luc/mg protein)

7
9
H2 H H NH3 NH3
N N N
X= N NH3 N NH3
H H S N N
N NH3
H NH3
Fig. 9 Comparative transfection efficiencies of structurally diverse thioureido-paCDs.
Fig. 10 Schematic representation of the formation of the supramolecular paCD-pDNA

nanocomplexes CDplexes and TEM image of a CDplex obtained from compound 18.
charged functionalities at secondary positions, might be beneficial for gene

vector design. However, the difficulties associated with secondary hydroxyl
groups reactivity, steric hindrance and positional isomerism make homo-
geneous functionalization at this face a far more complicated challenge. A
general strategy was developed starting from the facially differentiated
precursor 19, having the primary (O-6) hydroxyls protected as the corre-
sponding tert-dibutyldimethylsilyl ethers and the secondary (O-2 and O-3)
hydroxyls fully allylated. Hydroboration-oxidation (-20) generated 14
primary hydroxyls that were subjected to mesylation (-21) and nucleo-
philic displacement by Boc-protected cysteamine (-22). Sequential
hydrolysis of the silyl ether groups, acylation and Boc hydrolysis afforded a
first family of ‘‘reverse’’ cysteaminyl paCDs. Optimization of the transfec-
tion properties required the installation of myristoyl acyl groups at the
primary face (23, n=12) in this case. Alternatively, the cysteaminyl groups
Carbohydr. Chem., 2012, 38, 338–375 | 361

in the hexanoylated derivative (23, n=4) were transformed into ami-
noethylthioureido functionalities (24). Both 23 (n=12) and 24 (n=4)
condensed pDNA into monodisperse populations of very small-size cationic
CDplexes (47 15 nm, z-potential 39 1 mV and 40 3 nm, z-potential
44 1 mV, respectively, at N/P 10) that exhibited transfection efficiencies
similar to that of PEI-based polyplexes (Scheme 11).129
As an alternative approach for the construction of monodisperse poly-
cationic amphiphilic CDs, the archetypal click-type Cu(I)-catalyzed azide-
alkyne cycloaddition reaction was explored in collaboration with the group
of Santoyo-González. The bioorthogonal character of the chemical func-
tionalities at play and the unique property of CuAAC as a quantitative
ligation method make this methodology very well suited for the construc-
tion of vector libraries. The use of solid-supported Cu(I) catalysts130 was
found to be particularly well suited for multiple CuAAC with poly-
functionalized cyclodextrins while facilitating purification of the C7-sym-
metric macromolecular triazol adducts. The density of positive charges, their
distance to the CD core and the length of the lipophilic tails are key para-
meters that can be systematically modified to finely tune the transfection
capabilities. The results indicated that adjusting the hydrophilic/hydrophobic
n
OTBDMS 1. BBN OTBDMS O
2. NaOH, 1. TFA-CH2Cl2
O O OO
H2O2, 2. acylation
O O O
O 72% O O
O 7 O 7 n = 4, 82% O 7
n = 12, 34%
19 X S
X S
20, X = OH
88% H3N
21, X = OMs H3N
64% 22, X = S(CH2)2NHBoc 23
1. BocNH(CH2)2NCS
64% 2. TFA
O O O
O
O n
O O O O O
O O O
O O O OO
O O
O
O 7
S
S
OO O O O O O OO
O O O O O HN
S HN
HN S
HN
NH3
NH3
24
Scheme 11 Synthesis and schematic representation of ‘‘reverse’’ paCDs.
362 | Carbohydr. Chem., 2012, 38, 338–375

balance is critical to achieve optimal self-assembling properties and
stability of the resulting CDplexes.131 The benefits imparted by amphiphi-
licity appeared evident when comparing the behaviour within the branched
series 25, 26 and 27 (Fig. 11). Compound 26, bearing hexanoate tails, gave
rise to stable CDplexes that warranted full protection of pDNA from the
environment at N/P valuesZ5. Moreover, these CDplexes showed
remarkably high transfection efficiencies, analogous to that obtained using
Lipofectaminet 2000. Increasing hydrophobicity by replacing the hex-
anoate into tetradecanoate chains did not alter the complexing efficiency,
as seen from data for 27, but drastically decreased the transfection cap-
abilities, indicating that molecular modifications in paCDs do not only
affects the formation of CD-DNA nanoparticles, but also has dramatic
consequences on the further processes leading to cell internalization and
gene expression.
In general, triazole-type ‘‘click’’ paCDs were less efficient in promoting
gene transfection than thiourea-type derivatives in saline and serum-con-
taining media, but this could be corrected by implementing a ‘‘dual click’’
approach implying sequential CuAAC and thiourea-forming reactions (e.g.
28, Fig. 12).132 Broadening the spectrum of paCD structures available by
combining these two very robust methodologies opens new opportunities
for vector optimization.
Further work, carried out in collaboration with the group of Tros de
Ilarduya (Navarra, Spain), indicated that CDplexes formulated from a
plasmid codifying interleukin 12 (IL-12), a cytokine with anticancer activity,
and the dendritic aminoethylthioureido paCD derivative 29 were able to
transfect hepatocarcimoma Hep-G2 cells with significantly higher efficiency
than the corresponding polyplexes prepared with PEI.133 CDplexes loaded
with a luciferase-encoding reporter gene were next injected in a group of
mice to investigate the expression of the protein in different organs. The
nanocomplexes escaped from the lungs and exhibited a certain tropism to
the liver, thereby illustrating its potential in cellular hepatocarcinoma-
targeted gene therapies (Fig. 13).
NH3
+ + + N
+ NH3
+
+
++ N
N N
+
O
+ + RO
+ RO O
+
+ 7
25, R = H (non-amphiphilic)
26, R = hexanoyl
27, R = myristoyl
Fig. 11 Schematic representation of ‘‘click’’ paCDs and structures of some representative

examples.
Carbohydr. Chem., 2012, 38, 338–375 | 363

NH3
HN
ii) thiourea
HN formation
S
S
N NH3
N N
H H
i) CuAAC N
N N
O
O
O O
O
O 7
28
Fig. 12 Representative example of ‘‘dual click’’ paCD with highly efficient gene delivery
capabilities.
HN NH
N H N
S
HN NH
N
HN S 40
pg luciferase/mg protein
HN S
S O
O
O
O
O
35
N OO S
N
H
S
O
OO
O
O
O
30
HN
O O OO NH
pDNA 25
OO O O
O
O O H
N N 20
S
S O O O S 15
HN O O HN
S O O
NH O O O
OO O
O 10
O
N
O O
O S 5
S
NH
NH NH 0
S
NH S
N Heart Lung Spleen Liver
N NH
NH
HN 29
Fig. 13 Biodistribution of CDplexes obtained from paCD 29 after injection in mice.
An important issue in relationship with the possible preferential trans-

fection of certain cell lines is the elucidation of the routes by which
CDplexes are internalized and their fate inside the cell. This question has
been addressed in collaboration with the group of De Smedt (Ghent, Bel-
gium). The kinetics and the mechanism of internalization of CDplexes
formulated with paCD 30, bearing up to 21 amino groups and labelled with
the fluorescent probe lissamine-rhodamine, were investigated in kidney
epithelial VERO cells.134 The uptake of the CDplexes was monitored in real
time by confocal laser microscopy, revealing that this is a rather fast pro-
cess. The resulting endosomes tend to accumulate at the vicinity of the
nucleus after a certain time, pointing to the implication of active transport
mechanisms. These observations are consistent with internalization via
364 | Carbohydr. Chem., 2012, 38, 338–375

active-mediated endocytosis rather than through passive diffusion. Cla-
thrin-dependent and caveolin-dependent mechanisms were evidenced by
using inhibitors of the main internalization routes, but the endosomes
arising from the caveolae route were far more productive regarding trans-
fection (Fig. 14). It must be stressed, however, that those results only
applied strictly to the system under study. Other mechanisms may be
favoured for other paCDs or in other cell types. Thus, it has been recently
reported that macropinocytosis is the preferred internalization mechanism
for a cysteaminyl paCD derivative in intestinal epithelial Caco-2 cells.135
Since caveolae-mediated endocytosis and macropinocytosis operate in
many cell types, a main conclusion of the above work is that CDplexes
should be considered as broad range transfection systems. Although some
tissue selectivity can be achieved upon systemic injection, probably as a
function of particle size, cell-specific delivery would need the incorporation
of targeting ligands. Given the fundamental role of carbohydrates in cell
communication processes, the formulation of glycoCDplexes appeared
particularly attractive. Actually, glycoconjugation has been broadly used to
modulate the recognition and supramolecular properties of cyclodex-
trins,136 including amphiphilic CDs.137 In a first attempt, the combined use
of polycationinc amphiphilic CDs and neutral glycoamphiphilic CDs for
pDNA complexation was assayed. However, such formulations failed to
provide full DNA protection and mediate transfection, even when using low
NH3
N NH3
NH3
S
NH
N
NH S
S NH HN
O
NH S
O O
S
O SO3
O OO O
O n O N
O OO
O m
O
n+m=7 30 N
caveolae-mediated clathrin-dependent
pDNA
endocytosis (Cav-ME) endocytosis (CDE)
Gene expression
Fig. 14 Schematic representations of the internalization routes at play for CDplexes

formulated with the fluorescent paCD 30 in VERO cells.
Carbohydr. Chem., 2012, 38, 338–375 | 365

S S
H H H
N N NH3Cl N S
N N N
N N H H
H H S S HN
S OH
O
O O
O S NH O
O O O ClH3N OH
O O O
O O HO NH OH
O 7
7 HO
O
31 HO 32
HO
Fig. 15 Structures of the two prototypes of polycationic glycoamphiphilic cyclodextrins

(pGaCDs) 31 and 32.
(2%) proportions of the neutral glycocluster. It was reasoned that the

presence of neutral microdomains strongly destabilized the supramolecular
nanocomplexes. To overcome this problem, a new generation of poly-
cationic glycoamphiphilic CDs (pGaCDs), with regular arrangements of
glyco-cationic moieties, was designed.138 Two different prototypes were
considered, namely a derivative having the sugar moieties (mannosyl resi-
dues) and the cationic centres in separate branches (31) and another one in
which the cationic groups are located in the sugar moieties (32). In the
second case, the 6-amino-6-deoxyglucopyranosyl motif was chosen since it
is present in aminoglycoside antibiotics known to bind to the DNA double-
helix (Fig. 15).40
The synthesis of these macromolecules followed a convergent scheme in
which the heterobifunctional antenna, conveniently armed with a reactive
isothiocyanate group, was first prepared and coupled with the amphiphilic
cyclodextrin polyamine core 17 in a late step, as illustrated in Scheme 12 for
the mannosylated pGaCD 31. After the final deprotection reaction, the
target 7-fold symmetrical pGaCDs were obtained in homogeneous form, as
confirmed by MS, NMR and combustion analysis. Both compounds were
able to complex pDNA to form nanoparticulate glycoCDplexes that fully
protected the plasmid and promoted transfection of COS-7 cells.
The potential of the mannosylated pGaCD derivative for site specific
gene delivery was preliminary evaluated by measuring the ability of the
corresponding glycoCDplexes to bind the mannose specific lectin con-
canavalin A (Con A) by enzyme-linked lectin assay (ELLA). Further evi-
dence was obtained by developing a modified ELLA protocol for
recombinant human macrophage mannose receptor (rhMMR) and by
measuring the binding affinity towards murine macrophages in vitro.
CDplexes obtained from the non-glycosylated paCD or from the ami-
noglucosylated pGaCD were used as negative controls, whereas mannosy-
lated-BSA was employed in competitive assays to confirm the specific
mannose receptor-dependent character of the association. The results evi-
denced that mannosyl-CDplexes binding to macrophages can proceed
through non-specific electrostatic interactions with negatively charged
proteoglycans at the cell surface or through specific interaction with MMRs
(Fig. 16). At N/P 10, both mechanisms simultaneously operate. However, at
N/P 5 the MMR-dependent route is exceedingly dominant. CDplexes
366 | Carbohydr. Chem., 2012, 38, 338–375

OH OH
S O OH
NH2·HCl S
OH
O O
O BocHN N N
O H H
O N
O O + S
7 SCN
N N 33
70% 5 H H
17 H
N H
ClH3N N
N
S
O
OH NH
OH NH
O O
HN OH S OH
HN O HO OH ClH3N
OH OH
S
N
H NH 4
N H N
N S
4 NH NH
ClH3N HN
S S S S
NH NH NH
HN S
O 4
HN
HO OH S O O O
OH HN O
O OO O
OO S O
O O O O HO
O O OH O O S
O O OH
HO HO
S O
HN S
O
S HN
N N O O OO
HN 4 H H O O ClH3N
O O
HN S O
H S
O O S N H N
N O O OO O N NH
OO
S O O S NH
O O 4
NH3Cl NH
S S
HO
HO NH
S NH
HO
OH NH
O NH
S HO O O
NH
O
4
HO
HN S 4 S
HO OH
HN S HO OH N
HN H N
HO O H NH3Cl
HN HO N
S
N
O
NH3Cl N N
31 H H
Scheme 12 Synthesis of the pGaCD 31.
formulated at this N/P ratio specifically transfected macrophage RAW264.7

cells versus epithelial COS-7 or hepatic BNL-CL2 cells.138
Although the above commented results require validation in vivo, they
represent a proof of concept of the suitability of the strategy based on
pGaCDs to mimic the site-specific transfection capabilities of viruses. The
synthetic effort can be limited by statistic, but controlled, incorporation of
the appropriate glycotopes at the requested density. This strategy has been
already implemented for the preparation of galactosylated pGaCDs that
were able to complex pDNA and mRNA.139 The corresponding gly-
coCDplexes were internalized by hepatocytes via the asialoglycoprotein
receptor. Non-amphiphilic polycationic cyclodextrins exposing mannose,
Carbohydr. Chem., 2012, 38, 338–375 | 367

Fig. 16 Schematic representation of the competing cell internalization routes for glycoCD-
plexes. The lectin-mediated internalization pathway is largely favoured at N/P 5.
glucosamine and galactose residues have also been recently reported, but no
DNA complexing capabilities were studied.140
5 Outlook and perspectives

In this review, we covered the recent literature regarding the use of glyco-
materials in non-viral gene vector design. So far, most of the examples
exploit the potential of carbohydrates to improve the transfection cap-
abilities of preformed polymeric vectors by enhancing biodisponibility or
imparting targeting capabilities. But there are examples that demonstrate
that carbohydrates can play an active role in the interaction with poly-
nucleotides or in providing the right spatial orientation of key functional
elements for efficient DNA binding. Cyclodextrins are paradigmatic car-
bohydrate platforms for these channels. Their suitability to access mono-
disperse facial amphiphiles that self-associate in the presence of pDNA to
form nanocomplexes with transfectious capabilities has been already
demonstrated. Moreover, the nanometric cavity remains accessible in the
corresponding CDplexes, opening the door to nanoparticle surface coating
through supramolecular interactions. The CD commercial availability, easy
and relatively inexpensive synthesis of appropriate derivatives, facile pur-
ification, robustness and stability, biocompatibility, lack of immunogenicity
and safety match important criteria for non-viral vectors. Moreover, the
combination of a carbohydrate core and a carbohydrate coating in pGaCDs
offers new venues for the design of tailor-made artificial viruses. In all, there
should be plenty of room for carbohydrate chemists to develop high level
research connecting sugars and gene therapy.
Acknowledgments
The Spanish Ministerio de Economı́a y Competitividad (contract numbers
CTQ2010-15848 and SAF2010-15670; co-financed with the Fondo Europeo
de Desarrollo Regional FEDER) and the Junta de Andalucı́a are thanked
for funding.
368 | Carbohydr. Chem., 2012, 38, 338–375

References
1 N. Cartier, S. Hacein-Bey-Abina, C. C. Bartholomae, G. Veres, M. Schmidt,
I. Kutschera, M. Vidaud, U. Abel, L. Dal-Cortivo, L. Caccavelli, N.
Mahlaoui, V. Kiermer, D. Mittelstaedt, C. Bellesme, N. Lahlou, F. Lefrère,
S. Blanche, M. Audit, E. Payen, P. Leboulch, B. l’Homme, P. Bougnères, C.
Von Kalle, A. Fischer, M. Cavazzana-Calvo and P. Aubourg, Science, 2009,
326, 818.
2 S. Nayak and R. W. Herzog, Gene Ther., 2010, 17, 295.
3 C. C. Gao, Hum. Gene Ther., 2011, 22, A114.
4 J. K. Räty, J. T. Pikkarainen, T. Wirth and S. Ylä-Herttuala, Curr. Mol.
Pharmacol., 2008, 1, 13.
5 M. A. Mintzer and E. E. Simanek, Chem. Rev., 2009, 109, 259.
6 R. J. Christie, N. Nishiyama and K. Kataoka, Endocrinology, 2010, 151, 466.
7 M. E. Davis, Mol. Pharm., 2009, 6, 659.
8 D. J. Glover, H. J. Lipps and D. A. Jans, Nat. Rev. Genet., 2005, 6, 299.
9 D. Putnam, Nat. Mater., 2006, 5, 439.
10 B. Demeneix, Z. Hassani and J.-P. Behr, Curr. Gene Ther., 2004, 4, 445.
11 K. Itaka and K. Kataoka, Eur. J. Pharm. Biopharm., 2009, 71, 475.
12 J. H. Jeong, S. W. Kim and T. G. Park, Prog. Polym. Sci., 2007, 32, 1239.
13 G. F. Walker, C. Fella, J. Pelisek, J. Fahrmeir, S. Boeckle, M. Ogris and
E. Wagner, Mol. Ther., 2005, 11, 418.
14 H. Mok and T.-G. Park, Macromol. Biosci., 2009, 9, 731.
15 S. Ganta, H. Devalapally, A. Shahiwala and M. Amiji, J. Control. Release,
2008, 126, 187.
16 K. Knop, R. Hoogenboom, D. Fischer and U. S. Schubert, Angew. Chem. Int.
Ed., 2010, 49, 6288.
17 C. Biesa, C.-M. Lehra and J. F. Woodley., Adv. Drug Deliv. Rev., 2004, 56, 425.
18 H. Zhang, Y. Ma and X.-L. Sun, Med. Res. Rev., 2010, 30, 270.
19 S. Mao, W. Sun and T. Kissel, Adv. Drug Deliv. Rev., 2010, 62, 12.
20 For a recent survey on the use of naturally-occurring carbohydrate polymers
in gene delivery, see: A. Sizovs, P. M. McLendon, S. Srinivasachari and
T. M. Reineke, Top. Curr. Chem., 2010, 296, 131.
21 (a) L. Wightman, R. Kircheis, V. Rossler, S. Carotta, R. Ruzicka, M. Kursa
and E. Wagner, J. Gene Med., 2001, 3, 362; (b) Y. H. Choi, F. Liu, J.-S. Kim,
Y. K. Choi, J.-S. Park and S. W. Kim, J. Control. Release, 1998, 54, 39;
(c) P. Van de Wetering, E. E. Moret, N. M. E. Schuurmans-Nieuwenbroek,
M. J. Van Steenbergen and W. E. Hennink, Bioconjugate Chem., 1999, 10, 589;
(d) K. M. Kitchens, M. E. H. El-Sayed and H. Ghandehari, Adv. Drug Deliv.
Rev., 2005, 57, 2163.
22 S. Aoshima and S. Kanaoka, Chem. Rev., 2009, 109, 5245.
23 S. M. Moghimi, P. Symonds, J. C. Murray, A. C. Hunter, G. Debska and
A. Szewczyk, Mol. Ther., 2005, 11, 990.
24 T. Merdan, K. Kunath, H. Petersen, U. Bakowsky, K. H. Voigt, J. Kopecek
and T. Kissel, Bioconjugate Chem., 2005, 16, 785.
25 C. K. Chan and D. A. Jans, Gene Ther., 2001, 8, 166.
26 M. Kursa, G. F. Walker, V. Roessler, M. Ogris, W. Roedl, R. Kircheis and
E. Wagner, Bioconjugate Chem., 2003, 14, 222.
27 T. Kakudo, S. Chaki, S. Futaki, I. Nakase, K. Akaji, T. Kawakami,
K. Maruyama, H. Kamiya and H. Harashima, Biochemistry, 2004, 43, 5618.
28 (a) E. Wagner, C. Culmsee and S. Boeckle, Adv. Genet., 2005, 53, 333;
(b) J. Guo, K. A. Fisher, R. Darcy, C. F. Cryan and C. O’Driscoll, Mol.
Biosyst., 2010, 6, 1143.
Carbohydr. Chem., 2012, 38, 338–375 | 369

29 M. A. Zanta, P. Belguise-Valladier and J.-P. Behr, Proc. Natl Acad. Sci. USA,
1999, 96, 91.
30 (a) S. H. Kim, J. H. Jeong, K. C. Cho, S. W. Kim and T. G. Park, J. Control.
Release, 2005, 104, 223; (b) X. Zhao, H. Li and R. J. Lee, Expert Opin. Drug
Deliv., 2008, 5, 309.
31 (a) K. Kunath, T. Merdan, O. Hegener, H. Häberlein, T. Kissel, J. Gene Med.,
2003, 5, 588; (b) B. Lebleu, H. M. Moulton, R. Abes, G. D. Ivanova, S. Abes,
D. A. Stein, P. L. Iversen, A. A. Arzumanov and M. J. Gait, Adv. Drug Deliv.
Rev., 2008, 60, 517.
32 (a) M. Ogris, P. Steinlein, M. Kursa, K. Mechtler, R. Kircheis and E. Wagner,
Gene Ther., 1998, 5, 1425; (b) R. Huang, W. Ke, Y. Liu, C. Jiang and Y. Pei,
33 (a) B. Xu, S. Wiehle, J. A. Roth and R. J. Cristiano, Gene Ther., 1998, 5, 1235;
(b) H. Lee, T. H. Kim and T. G. Park, J. Control. Release, 2002, 83, 109.
34 (a) M. C. Palanca-Wessels, A. J. Convertine, R. Cutler-Strom, G. C. Booth, F.
Lee, G. Y. Berguig, P. S. Stayton and O. W. Press, Mol. Ther., 2011, 19, 1529;
and (b) V. Torchilin, Expert Opin. Drug Deliv., 2008, 5, 1003.
35 (a) G. Y. Wu, C. H. Wu, J. Biol. Chem., 1987, 262, 4429; (b) G. Y. Wu and
C. H. Wu, J. Biol. Chem., 1988, 263, 14621
36 S. C. De Smedt, J. Demeester and W. E. Hennink, Pharm. Res., 2000, 17, 113.
37 S. Höbel, A. Loos, D. Appelhans, S. Schwarz, J. Seidel, B. Voit and A. Aigner,
J. Control. Release, 2011, 149, 146.
38 (a) S. P. Strand, M. M. Issa, B. E. Christensen, K. M. Vårum and P.
Artursson, Biomacromolecules, 2008, 9, 3268; (b) S. P. Strand, S. Lelu, N. K.
Reitan, C. de Lange Davies, P. Artursson and K. M. Vårum, Biomaterials,
2010, 31, 975.
39 K. Koumoto, M. Mizu, T. Anada, T. Nagasaki, S. Shinkai and K. Sakurai,
Bull. Chem. Soc. Jpn., 2005, 78, 1821.
40 (a) L. Desigaux, M. Sainlos, O. Lambert, R. Chevre, E. Letrou-Bonneval, J.-P.
Vigneron, P. Lehn, J.-M. Lehn and B. Pitard, Proc. Natl. Acad. Sci. USA,
2007, 104, 16534; (b) M. Sainlos, M. Hauchecorne, N. Oudrhiri, S. Zertal-
Zidani, A. Aissaoui, J.-P. Vigneron, J.-M. Lehn and P. Lehn, ChemBioChem,
2005, 6, 1023.
41 (a) H. Huang, G. Tang, Q. Wang, D. Li, F. Shen, J. Zhoub and Hai Yu, Chem.
Commun., 2006, 2382; (b) G. P. Tang, H. Y. Guo, F. Alexis, X. Wang, S. Zeng,
T. M. Lim, J. Ding, Y. Y. Yang, S. Wang, J. Gene Med., 2006, 8, 736.
42 (a) H. Huang, H. Yu, G. Tang, Q. Wang and J. Li, Biomaterials, 2010, 31,
1830; (b) Q.-Y. Jiang, L.-H. Lai, J. Shen, Q.-Q. Wang, F.-J. Xu and G.-P.
Tang, Biomaterials, 2011, 32, 7253.
43 Y. Ping, C. Liu, Z. Zhang, K. L. Liu, J. Chen and J. Li, Biomaterials, 2011, 32,
8328.
44 X. Liu, P. Rocchi and L. Peng, New J. Chem., 2012, 36, 256.
45 (a) H. Arima, F. Kihara, F. Hirayama and K. Uekama, Bioconjugate Chem.,
2001, 12, 476; (b) F. Kihara, H. Arima, T. Tsutsumi, F. Hirayama and K.
Uekama, Bioconjugate Chem., 2002, 13, 1211.
46 T. Tsutsumi, F. Hirayama, K. Uekama and H. Arima, J. Control. Release,
2007, 119, 349.
47 (a) H. Arima, Y. Chihara, M. Arizono, S. Yamashita, K. Wada, F. Hirayama
and K. Uekama, J. Control. Release, 2006, 116, 64; (b) Y. Chihara, H. Arima,
M. Arizono, K. Wada, F. Hirayama and K. Uekama, J. Incl. Phenom. Macro-
cycl. Chem., 2006, 56, 89.
48 (a) M. Hashimoto, M. Morimoto, H. Saimoto, Y. Shigemasa and T. Sato,
Bioconjugate Chem., 2006, 17, 309; (b) M. Hashimoto, M. Morimoto,
370 | Carbohydr. Chem., 2012, 38, 338–375

H. Saimoto, Y. Shigemasa, H. Yanagie, M. Eriguchi and T. Sato, Biotechnol.
Lett., 2006, 28, 815.
49 M. Ahmed, X. Jiang, Z. Deng and R. Narain, Bioconjugate Chem., 2009, 20,
2017.
50 M. Ahmed, Z. Deng and R. Narain, ACS Appl. Mater. Interfaces, 2009, 9,
1980.
51 J.-M. Li, M.-X. Zhao, H. Su, Y.-Y. Wang, C.-P. Tan, L.-N. Ji and Z.-W. Mao,
52 (a) J. Li and X. J. Loh, Adv. Drug. Deliv. Rev., 2008, 60, 1000; (b) J. J. Li, F.
Zhao and J. Li, Appl. Microbiol. Biotechnol., 2011, 90, 427.
53 (a) I.-K. Park, H. A. von Recum, S. Jiang and S. H. Pun, Langmuir 2006, 22,
8478; (b) X. Zhang, K. K. Sharma, M. Boeglin, J. Ogier, D. Mainard, J.-C.
Voegel, Y. Mély and N. Benkirane-Jessel, Nano Lett., 2008, 8, 2432.
54 S.-J. Kim, H. Ise, M. Goto, K. Komura, C.-S. Cho and T. Akaike, Bioma-
terials, 2011, 32, 3471.
55 D. B. Rozema, D. L. Lewis, D. H. Wakefield, S. C. Wong, J. J. Klein, P. L.
Roesch, S. L. Bertin, T. W. Reppen, Q. Chu, A. V. Blokhin, J. E. Hagstrom
and J. A. Wolff, Proc. Natl. Acad. Sci USA, 2007, 104, 12982.
56 (a) M.-A. Zanta, O. Boussif, A. Adib and J.-P. Behr, Bioconjugate Chem.,
1997, 8, 839; (b) S. S. Diebold, M. Kursa, E. Wagner, M. Cotten and M.
Zenke, J. Biol. Chem., 1999, 274, 19087; (c) T. Bettinger, J.-S. Remy and P.
Erbacher, Bioconjugate Chem., 1999, 10, 558.
57 (a) P. Erbacher, A. C. Roche, M. Monsigny and P. Midoux, Bioconjugate
Chem., 1995, 6, 401; (b) J. C. Perales, T. Ferkol, H. Beegen, O. D. Ratnoff and
R. W. Hanson, Proc. Natl. Acad. Sci. USA 1994, 91, 4086; (c) T. Ferkol, J. C.
Perales, F. Mularo and R. W. Hanson, Proc. Natl. Acad. Sci. USA 1996, 93,
101.
58 H. Arima, S. Yamashita, Y. Mori, Y. Hayashi, K. Motoyama, K. Hattori, T.
Takeuchi, H. Jono, Y. Ando, F. Hirayama and K. Uekama, J. Control.
Release, 2010, 146, 106.
59 (a) K. Motoyama, Y. Nakashima, Y. Aramaki, F. Hirayama, K. Uekama and
H. Arima, J. Drug Deliv., 2011, Article ID 476137, doi:10.1155/2011/476137;
(b) T. Montier, P. Delépine, T. Benvegnu, V. Ferrières, M.-L. Miramon, S.
Dagorn, C. Guillaume, D. Plusquellec and C. Férec, Blood Cells Mol. Dis.
2004, 32, 271; (c) M. Hashida, S. Kawakami and F. Yamashita, Chem. Pharm.
Bull., 2005, 53, 871.
60 C. Rondanino, M. T. Bousser, M. Monsigny and A. C. Roche, Glycobiology,
2003, 13, 509.
61 (a) A. C. Roche, I. Fajac , S. Grosse, N. Frison, C. Rondanino, R. Mayer and
M. Monsigny, Cell. Mol. Life Sci., 2003, 60, 288; (b) M. Monsigny, C.
Rondanino, E. Duverger, I. Fajac and A.-C. Roche, Biochim. Biophys. Acta,
2004, 1673, 94.
62 (a) S. Grosse, A. Tremeau-Bravard, Y. Aron, P. Briand, I. Fajac, Gene Ther.,
2002, 9, 1000; (b) M. Hashimoto, M. Morimoto, H. Saimoto, Y. Shigemasa
and T. Sato, Bioconjugate Chem., 2006, 17, 309.
63 S. Grosse, G. Thévenot, M. Monsigny and I. Fajac, J. Gene Med., 2006, 8, 845.
64 (a) H. Akita, T. Masuda, T. Nishio, K. Niikura, K. Ijiro and H. Harashima,
Mol. Pharm., 2011, 8, 1436; (b) T. Masuda, H. Akita, T. Nishio, K. Niikura,
K. Kogure, K. Ijiro and H. Harashima, Biomaterials, 2008, 29, 709.
65 M. Breton, J.-F. Berret, C. Bourgaux, T. Kral, M. Hof, C. Pichon, M.
Bessodes, D. Scherman and N. Mignet, Langmuir, 2011, 27, 12336.
66 P. Vicennati, A. Giuliano, G. Ortaggi and A. Masotti, Curr. Med. Chem.,
2008, 15, 2826.
Carbohydr. Chem., 2012, 38, 338–375 | 371

67 M. Huang, C.-W. Fong, E. Khor and L.-Y. Lim, J. Control. Release, 1999,
106, 391.
68 C. Troiber and E. Wagner, Bioconjugate Chem., 2011, 22, 1737.
69 C. Lin and J. F. J. Engbersen, J. Control. Release, 2008, 132, 267.
70 Y. Liu, L. Wenning, M. Lynch and T. M. Reineke, J. Am. Chem. Soc., 2004,
126, 7422.
71 M. Metzke, N. O’Connor, S. Maiti, E. Nelson and Z. Guan, Angew. Chem.
Int. Ed., 2005, 44, 6529.
72 M. Metzke and Z. Guan, Biomacromolecules, 2008, 9, 208.
73 N. P. Ingle, B. Malone and T. M. Reineke, Trends Biotechnol., 2011, 29, 443.
74 (a) Y. Liu and T. M. Reineke, J. Am. Chem. Soc. 2005, 127, 3004; (b) Y. Liu
and T. M. Reineke, Bioconjugate Chem., 2006, 17, 101.
75 Y. Liu and T. M. Reineke, Biomacromolecules, 2010, 11, 316.
76 C.-C. Lee, Y. Liu and Theresa M. Reineke, Bioconjugate Chem., 2008, 18, 428.
77 Y. Liu and T. M. Reineke, Bioconjugate Chem., 2007, 18, 19.
78 http://www.techulon.com/products, last visited March 2012.
79 T. M. Reineke and M. E. Davis, Bioconjugate Chem., 2003, 14, 247.
80 (a) S. Srinivasachari, Y. Liu, G. Zhang, L. Prevette and T. M. Reineke, J. Am.
Chem. Soc., 2006, 125, 8176; (b) S. Srinivasachari, Y. Liu, L. Prevette and
T. M. Reineke, Biomaterials, 2007, 28, 2885.
81 R. D. Lins, C. S. Pereira and P. H. Hünenberger, Proteins, 2004, 55, 177.
82 M. Meldal and C. W. Tornøe, Chem. Rev., 2008, 108, 2952.
83 M. E. Davis and M. E. Brewster, Nat. Rev. Drug Discov., 2004, 3, 1023.
84 J. E. Zuckerman, C. H. J. Choi, H. Han and M. E. Davis, Proc. Natl. Acad.
Sci. USA, 2012, 109, 3137.
85 H. Gonzalez, S. J. Hwang and M. E. Davis, Bioconjugate Chem., 1999, 10,
1068.
86 S. J. Hwang, N. C. Bellocq and M. E. Davis, Bioconjugate Chem., 2001, 12, 280.
89 S. R. Popielarski, S. Mishra and M. E. Davis, Bioconjugate Chem., 2003
14, 672.
90 M. E. Davis and N. Q. Bellocq, J. Incl. Phenom. Macrocycl. Chem., 2003
44, 17.
91 M. E. Davis, S. H. Pun, N. C. Bellocq, T. M. Reineke, S. R. Popielarski, S.
Mishra and J. D. Heidel, Curr. Med. Chem., 2004, 11, 179.
92 S. Mishra, J. D. Heidel, P. Webster and M. E. Davis, J. Control. Release, 2006,
116, 179.
93 J.-P. Behr, Chimia, 1997, 51, 34.
94 S. Srinivasachari and T. M. Reineke, Biomaterials, 2009, 30, 928.
95 S. Mishra, P. Webster and M. E. Davis, Eur. J. Cell Biol., 2004, 83, 97.
96 S. H. Pun and M. E. Davis, Bioconjugate Chem., 2002, 13, 630.
97 N. C. Bellocq, S. H. Pun, G. S. Jensen and M. E. Davis, Bioconjugate Chem.,
2003, 14, 1122.
98 D. W. Bartlett and M. E. Davis, Bioconjugate Chem., 2007, 18, 456.
99 S. H. Pun, F. Tack, C. C. Bellocq, J. Cheng, B. H. Grubbs, G. S. Jensen, M. E.
Davis, M. Brewster, M. Janicot, B. Janssens, W. Floren and A. Bakker,
Cancer Biol. Ther., 2004, 3, 641.
100 S. Hu-Lieskovan, J. D. Heidel, D. W. Bartlett, M. E. Davis and T. J. Triche,
Cancer Res., 2005, 65, 8984.
101 D. W. Bartlett and M. E. Davis, Biotechnol. Bioeng., 2008, 99, 975.
102 J. D. Heidel, Z. Yu, J. Y. Liu, S. M. Rele, Y. Liang, R. K. Zeidan, D. J.
Kornbrust and M. E. Davis, Proc. Natl. Acad. Sci. USA, 2007, 104, 5715.
372 | Carbohydr. Chem., 2012, 38, 338–375

103 http://clinicaltrials.gov/ct2/show/NCT00689065
104 M. E. Davis, J. E. Zuckerman, C. H. J. Choi, D. Seligson, A. Tolcher, C. A.
Alabi, Y. Yen, J. D. Heidel and A. Ribas, Nature, 2010, 464, 1067.
105 For the most recent updates, see: http://calando.interactive.biz/technology/
rondel/ or http://clinicaltrials.gov/ct2/show/NCT00689065.
106 (a) B. A. Lowe and R. Wang, Polymer, 2007, 48, 2221; (b) L. Albertin, M.
Stenzel, C. Barner-Kowollik, L. John, R. Foster and T. P. Davis, Macro-
molecules, 2004, 37, 7530; (c) L. Liu, J. Zhang, W. Lv, Y. Luo and X. Wang,
J. Polym. Sci. Part A: Polym. Chem., 2010, 48, 3350; (d) C. Boyer and T. P.
Davis, Chem. Commun., 2009, 6029.
107 M. Ahmed and R. Narain, Biomaterials, 2011, 32, 5279.
108 A. E. Smith, A. Sizovs, G. Grandinetti, L. Xue and T. M. Reineke, Bioma-
cromolecules, 2011, 12, 3015.
109 T. Fujiwara, S. Hasegawa, N. Hirashima, M. Nakanishi and T. Ohwada,
Biochim. Biophys. Acta-Biomembr., 2000, 1468, 396.
110 C. Ortiz Mellet, J. M. Benito and J. M. Garcı́a Fernández, Chem. Eur. J., 2010,
16, 6728.
111 (a) C. Li, H. Tian, N. Rong, K. Liu, F. Liu, Y. Zhu, R. Qiao and Y. Jiang,
Biomacromolecules, 2011, 12, 298; (b) H. Yan, Z.-F. Li, Z.-F. Guo, Z.-L. Lu,
F. Wang and L.-Z. Wu, Bioorg. Med. Chem., 2012, 20, 801; (c) Q.-D. Huang,
H. Chen, L.-H. Zhou, J. Huang, J. Wu and X.-Q. Yu, Chem. Biol. Drug Des.,
2008, 71, 224; (d) L.-H. Zhou, M. Yang, H. Zhou, J. Zhang, K. Li, Y.-Z.
Xiang, N. Wang, Y.-F. Tian and X.-Q. Yu, Chem. Biol. Drug Des., 2009, 73,
216; (e) J.-L. Liu, Q.-P. Ma, Q.-D. Huang, W.-H. Yang, J. Zhang, J.-Y. Wang,
W. Zhu and X.-Q. Yu, Eur. J. Med. Chem., 2011, 46, 4133; (f) W.-J. Yi, Z.-H.
Feng, Q.-F. Zhang, J. Zhang, L.-D. Li, W. Zhu and X.-Q. Yu, Org. Biomol.
Chem., 2011, 9, 2413; (g) Q.-D. Huang, G.-X. Zhong, Y. Zhang, J. Ren, Y. Fu,
J. Zhang, W. Zhu and X.-Q. Yu, PLoS ONE, 2011, 6, e23134.
112 (a) F. Sansone, M. Dudič, G. Donofrio, C. Rivetti, L. Baldini, A. Casnati, S.
Cellai and R. Ungaro, J. Am. Chem. Soc., 2006, 128, 14528; (b) R. V. Rodik,
A. S. Klymchenko, N. Jain, S. I. Miroshnichenko, L. Richert, V. I. Kalchenko
and Y. Mély, Chem. Eur. J., 2011, 17, 5526.
113 (a) Y. Aoyama, Bull. Chem. Soc. Jpn., 2009, 82, 419; (b) S. Horiuchi and Y.
Aoyama, J. Control. Release, 2006, 116, 107.
114 (a) D. A. Fulton and J. F. Stoddart, Bioconjugate Chem., 2001, 12, 655; (b) Y.
M. Chabre and R. Roy, Adv. Carbohydr. Chem. Biochem., 2010, 63, 165.
115 (a) Y. Aoyama, Chem. Eur. J. 2004, 10, 588; (b) Y. Aoyama, Trends
Glycosci. Glycotechnol., 2005, 17, 39; (c) Y. Aoyama, Chem. Eur. J., 2009, 82, 419.
116 Y. Aoyama, Y. Tanaka and S. Sugahara, J. Am. Chem. Soc., 1989, 111, 5397.
117 O. Hayashida, K. Mizuki, K. Akagi, A. Matsuo, T. Kanamori, T. Nakai, S.
Sando and Y. Aoyama, J. Am. Chem. Soc., 2003, 125, 594.
118 (a) M. Lahmann, Top. Curr. Chem., 2009, 288, 17; (b) C. Ortiz Mellet, J.
Defaye and J. M. Garcı́a Fernández, Chem. Eur. J., 2002, 8, 1982.
119 (a) E. M. Nestorovich, V. A. Karginov, M. R. Popoff, S. M. Bezrukov and H.
Barth, PLoS ONE 2011, 6, e23927; (b) A. Dı́az-Moscoso, A. Méndez-Ardoy,
F. Ortega-Caballero, J. M. Benito, C. Ortiz Mellet, J. Defaye, T. M. Robinson,
A. Yohannes, V. A. Karginov and J. M. Garcı́a Fernández, ChemMedChem,
2011, 6, 181; (c) V. A. Karginov, E. M. Nestorovich, M. Moayeri, S. H. Leppla
and S. M. Bezrukov, Proc. Natl. Acad. Sci. USA, 2005, 102, 15075.
120 C. Ortiz Mellet, J. M. Garcı́a Fernández and J. M. Benito, Chem. Soc. Rev.,
2011, 40, 1586.
121 S.-A. Cryan, A. Holohan, R. Donohue, R. Darcy and C. M. O’Driscoll, Eur.
J. Pharm. Sci., 2004, 21, 625.
Carbohydr. Chem., 2012, 38, 338–375 | 373

122 (a) N. Mourtzis, K. Eliadou, C. Aggelidou, V. Sophianopoulou, I. M. Mavridis
and K. Yannakopoulou, Org. Biomol. Chem., 2007, 5, 125; (b) N. Mourtzis, M.
Paravatou, I. M. Mavridis, M. L. Roberts and K. Yannakopoulou, Chem. Eur.
J., 2008, 14, 4188.
123 S. Srinivasachari, K. M. Fichter and T. M. Reineke, J. Am. Chem. Soc., 2008,
130, 4618.
124 V. Bennevault-Celton, A. Urbach, O. Martin, C. Pichon, P. Guégan and P.
Midoux, Bioconjugate Chem., 2011, 22, 2404.
125 (a) S.-A. Cryan, R. Donohue, B. J. Ravoo, R. Darcy, C. M. O’Driscoll,
J. Drug. Deliv. Sci. Technol., 2004, 14, 57; (b) A. McMahon, E. Gomez,
R. Donohue, D. Forde, R. Darcy and C. M. O’Driscoll, J. Drug. Deliv. Sci.
Technol., 2008, 18, 303
126 A. Dı́az-Moscoso, P. Balbuena, M. Gómez-Garcı́a, C. Ortiz Mellet, J. M.
Benito, L. Le Gourriérec, C. Di Giorgio, P. Vierling, A. Mazzaglia, N. Micali,
J. Defaye and J. M. Garcı́a Fernández, Chem. Commun., 2008, 2001.
127 (a) A. K. H. Hirsch, F. R. Fischer and F. Diederich, Angew. Chem. Int. Ed.,
2007, 46, 338; (b) E. A. Katayev, Y. A. Ustynyuk and J. L. Sessler, Coord.
Chem. Rev., 2006, 250, 3004.
128 A. Dı́az-Moscoso, L. Le Gourriérec, M. Gómez-Garcı́a, J. M. Benito,
P. Balbuena, F. Ortega-Caballero, N. Guilloteau, C. Di Giorgio, P. Vierling,
J. Defaye, C. Ortiz Mellet and J. M. Garcı́a Fernández, Chem. Eur. J., 2009,
15, 12871.
129 F. Ortega-Caballero, C. Ortiz Mellet, L. Le Gourriérec, N. Guilloteau, C. Di
Giorgio, P. Vierling, J. Defaye and J. M. Garcı́a Fernández, Org. Lett., 2008,
10, 5143.
130 A. Megia-Fernandez, M. Ortega-Muñoz, J. Lopez-Jaramillo, F. Hernandez-
Mateo and F. Santoyo-Gonzalez, Adv. Synth. Catal., 2010, 352, 3306.
131 A. Méndez-Ardoy, M. Gómez-Garcı́a, C. Ortiz Mellet, N. Sevillano, M. D.
Girón, R. Salto, F. Santoyo-González and J. M. Garcı́a Fernández, Org.
Biomol. Chem., 2009, 7, 2681.
132 A. Méndez-Ardoy, N. Guilloteau, C. Di Giorgio, P. Vierling, F. Santoyo-
Gonzalez, C. Ortiz Mellet and J. M. Garcı́a Fernández, J. Org. Chem., 2011,
76, 5882.
133 A. Méndez-Ardoy, K. Urbiola, C. Aranda, C. Ortiz-Mellet, J. Manuel Garcı́a-
Fernández and C. Tros de Ilarduya, Nanomedicine, 2011, 6, 1697.
134 A. Dı́az-Moscoso, D. Vercauteren, J. Rejman, J. M. Benito, C. Ortiz Mellet,
S. C. de Smedt and J. M. Garcı́a Fernández, J. Control. Release, 2010, 143,
318.
135 (a) M. J. O’Neil, J. Guo, C. Byrne, R. Darcy and C. M. O’Driscoll, Int.
J. Pharm., 2011, 413, 174; (b) C. Byrne, F. Sallas, D. K. Rai, J. Ogier and
R. Darcy, Org. Biomol. Chem., 2009, 7, 3763.
136 (a) M. Gómez-Garcı́a, J. M. Benito, R. Gutiérrez-Gallego, A. Maestre,
C. Ortiz Mellet, J. M. Garcı́a Fernández and J. L. Jiménez Blanco, Org.
Biomol. Chem., 2010, 8, 1849; (b) M. Gómez-Garcı́a, J. M. Benito, A. P.
Butera, C. Ortiz Mellet, J. M. Garcı́a Fernández and J. L. Jiménez Blanco,
J. Org. Chem., 2012, 77, 1273; (c) J. M. Benito, M. Gómez-Garcı́a, C. Ortiz
Mellet, I. Baussanne, J. Defaye and J. M. Garcı́a Fernández, J. Am. Chem.
Soc., 2004, 126, 10355; (d) D. Grünstein, M. Maglinao, R. Kikkeri, M. Collot,
K. Barylyuk, B. Lepenies, F. Kamena, R. Zenobi and P. H. Seeberger, J. Am.
Chem. Soc. 2011, 133, 13957; (e) A. Vargas-Berenguel, F. Ortega-Caballero
and J. M. Casas-Solvas, Mini-Rev. Org. Chem., 2007, 4, 1.
137 (a) R. V. Vico, J. Voskuhl and B. J. Ravoo, Langmuir, 2011, 27, 1391;
(b) J. Voskuhl, M. C. A. Stuart and B. J. Ravoo, Chem. Eur. J., 2010, 16, 2790;
374 | Carbohydr. Chem., 2012, 38, 338–375

(c) S. McNicholas, A. Rencurosi, L. Lay, A. Mazzaglia, L. Sturiale, M. Perez
and R. Darcy, Biomacromolecules, 2007, 8, 1851.
138 A. Dı́az-Moscoso, N. Guilloteau, C. Bienvenu, A. Méndez-Ardoy, J. L.
Jiménez Blanco, J. M. Benito, L. Le Gourriérec, C. Di Giorgio, P. Vierling,
J. Defaye, C. Ortiz Mellet and J. M. Garcı́a Fernández, Biomaterials, 2011, 32,
7263.
139 (a) N. Symens, A. Méndez-Ardoy, J. M. Garcı́a Fernández, S. C. De Smedt,
and J. Rejman, Human Gene Ther., 2011, 22, A122; (b) N. Symens, A. Méndez-
Ardoy, A. Dı́az-Moscoso, E. Sánchez-Fernández, K. Remaut, J. Demeester, J.
M. Garcı́a Fernández, S. C. De Smedt and J. Rejman, Bioconjugate Chem.,
2012, DOI: 10.10217bc3001003.
140 M. Lampropoulou and K. Yannakopoulou, J. Incl. Phenom. Macrocycl.
Chem., 2011, 70, 345.
Carbohydr. Chem., 2012, 38, 338–375 | 375

Furanose-based templates in the
chemoselective generation of molecular
diversity
Ana M. Gómez,* Clara Uriel and J. Cristóbal López*
DOI: 10.1039/9781849734769-00376
Highly functionalized 1-exo-alkylidene-2,3-anhydro-, and 1’-halo-1-exo-

alkylidene-2,3-anhydro-furanoses, which are accessible from D-mannose in
four or five steps, respectively, can be efficiently transformed in furanose
based templates with up to four sites of molecular diversity, by way of a
series of chemoselective transformations at their double bond, oxirane, or
allylic oxirane, functionalities.
1 Introduction
One of the hurdles associated to drug discovery is related to the spatial (3D)
display of pharmacophoric groups to attain a desired biological response.1
In this context, carbohydrates were soon recognized as privileged platforms
for appending pharmacophores with stereo-determined orientations.2
Monosaccharides are endowed with rigid core-like characteristics and are
available at low cost with a variety of spatial orientations of their hydroxyl
substituents. They can also enjoy pyranose or furanose structures, which
make them excellent platforms to tailor molecular diversity by appending
desired substituents at selected positions. An example is shown in Fig. 1,
where a b-D-glucopyranoside with five potential pharmacophoric units, one
at each of its five functionalized positions, is displayed.
Based on the concept that a bioactive molecule consists on a display of
pharmacophoric units, responsible for binding, assembled around an inert,
non-binding, structure acting as a scaffold,3 Hirschmann, Nicolaou, Smith
and co-workers developed, in 1992, a peptidomimetic (2, Fig. 2) that was
able to interact with somatostatin receptors.4 In their approach, a b-D-
glucopyranose structure functioned as the scaffold to which several phar-
macophores were attached. In their landmark work, several b-D-glucopyr-
anose derivatives were shown to mimic the active conformation of the cyclic
hexapeptide L-363,301 (1), a potent somatostatin receptor agonist. Thus,
the sugar template replaces the metabolically labile peptide backbone and
serves to control the relative positioning and orientation of the binding side
chains.
The usefulness of sugar scaffolds as peptide backbone surrogates con-
tinued to be exploited by Kessler and co-workers that developed several
peptidomimetics, based on b-D-mannopyranose scaffolds, for the integrin
family.5 Even though most examples of sugar-based scaffolds as peptido-
mimetics are based on pyranosidic structures, furanoses have also been
studied by Hanessian et al.6 They described furanose-based scaffolds as
Instituto de Quı´mica Orgánica General, IQOG-CSIC, Juan de la Cierva 3, 28006 Madrid,

Spain. E-mail: ana.gomez@csic.es (AMG); jc.lopez@csic.es (JCL)
376 | Carbohydr. Chem., 2012, 38, 376–397

O
O
Fig. 1 Spatial arrangement of substituents on a b-D-glucopyranose core.
OH NH2
O NH2
H
N N
O O O
H
O NH
N O
O BnO O
H
N N
H NH O O
O NH
1 2
Fig. 2 Hexapeptide L-363,301 (1), a potent somatostatin receptor agonist, and a b-D-gluco-
pyranose-based peptidomimetic, 2.
conformationally constrained analogs of acyclic succinic acid-based inhi-

bitors of metalloproteinases.
The successful use of carbohydrates as scaffolds in the area of peptido-
mimetics triggered a considerable research effort on the use of sugar deri-
vatives as templates in bioactive compound discovery.7 Carbohydrate
templates have consequently been generated from pyranoses,8 furanoses,9
disaccharides10 and bicyclic sugar derivatives.11 Sugar scaffolds also provide
an unparalleled opportunity to generate libraries of high functional and
structural diversity, with most of the strategies for their construction having
been based on the stepwise derivatization of orthogonally protected carbo-
hydrate derivatives.12–14
2 Design and synthesis of functionalized furanose-based precursors

Based on these literature precedents, we were interested in the design of less-
frequented furanose-based templates by chemoselective incorporation of
the sought pharmacophoric units on a functionalized furanose-core, rather
than a stepwise incorporation of appendages onto an orthogonally differ-
entiated furanose polyol. Our furanose precursor should then be endowed
with functionalities other than the current ‘‘hydroxyl groups’’. Additionally,
in order to make this template appealing, our furanose scaffold should be
accessible in few efficient synthetic steps from a readily available starting
material.
2.1 Epoxy exo-glycals as polyfunctionalized precursors for furanose-based

templates
Along these lines, we recently embarked on a project aimed at the synthesis
of furanose-based templates 3, from polyfunctionalized furanose derivatives
4a–c (Scheme 1). Compound 4a can be subjected to simple chemical
Carbohydr. Chem., 2012, 38, 376–397 | 377

acetonide
olefin
R 1 O 6
O X
O alkenyl
O
O
O 5 1 1’ halide (X = Br, I)
3
allylic
O oxirane
3
RO
NR2 oxirane
3 4
a) X = H; b) X = Br; c) X = I
Scheme 1 Functionalized furanose-based precursors 4 for the preparation of furanose

templates 3.
reactions that would lead to diversified platforms, e.g. 3, in which the three
different substituents could be incorporated sequentially, and in a com-
pletely stereocontrolled manner.
In this context, compound 4a contains an allylic oxirane moiety, which
could be dissected (reactivity-wise) into an oxirane and an exocyclic eno-
lether-type olefin. In derivatives 4b,c, the exocyclic double bond is further
functionalized with the presence of a halogen atom. Particularly appealing
to us, was that furanoses 4 could display different sets of reactivity, e.g.
towards nucleophiles and/or electrophiles. For instance, reaction at the
exocyclic olefin in compound 4a would still leave the oxirane moiety for
future transformations, likewise reaction of the oxirane moiety in 4a would
leave the enol ether functionality intact. Besides this postulated reaction
pattern, compound 4a could also display allylic-oxirane reactivity. Fur-
thermore, the presence of a halogen atom in furanoses 4b or 4c would still
add a supplementary handle for further functionalization at C-1’ in these
derivatives.
2.2 Synthesis of epoxy-exo glycal 4a, from D-mannose

Our synthetic protocol for the synthesis of compounds 4 took advantage of
some transformations described in our laboratory. Thus, reaction of fur-
anosyl chloride 7 with MeLi provided a good route to C-1 methyl glycal 8
(Scheme 2b).15 Access to glycosyl chloride 7 could be done by reaction of
either thermodynamic, i.e. 5,16 or kinetic, i.e. 6,17 mannose-diacetonides
with triphenylphosphane in carbon tetrachloride (Scheme 2a).18
The desired epoxy-exo glycal 4a was readily obtained from C-1 methyl
glycal 8 in just one synthetic operation by treatment with bromine and
triethylamine in dichloromethane (8-4a, Scheme 3).19 This one-pot
transformation involves three widely-known transformations in carbohy-
drate chemistry, namely: i) the Lemieux, Fraser-Reid electrophilic
bromination of glycals,20 ii) the base-mediated formation of oxiranes from
1,2-halohydrins,21 and iii) the elimination of anomeric halides mediated by
base.22 Although we have no proof of the actual steps order in the overall
transformation, in our opinion, the sequence will likely start with the ste-
reoselective bromination of the electron rich double bond (step i, Scheme 3)
to furnish dibromides 9, and since no reaction intermediate (10 or 11) has
ever been observed (TLC) in these transformations, a pathway 9-11-4a
(involving a reactive allylic bromide) seems more likely.
378 | Carbohydr. Chem., 2012, 38, 376–397

O
O O O OH
PPh3, CCl4
P2O5 O O THF
O
95%
90%
O Cl
5 O
(a) D-mannose
OMe
O O
PPh3, CCl4
p-TsOH O
O THF
sikkon O
O 7
O 95%
DMF
85% OH
6
O O
O Cl MeLi,
O O O
then NH4Cl
(b) THF, –78 °C

O O HO
74%
7 8
Scheme 2 Synthesis of C-1 methyl glycal 8, via furanosyl chloride 7, from D-mannose.
O
O CH3
O
Br
O ii) iii)
O
O O CH3 10
Br2
8 O 4a
NEt3 Br
i) HO Br O
O CH2
9 iii) ii)
HO Br
11
Steps: i) halogenation; ii) epoxide formation; iii) elimination
Scheme 3 One-pot transformation of furanosidic C-1 methyl glycal 8 into epoxy-exo-glycal 4a.
2.3 Synthesis of epoxy-1’-halo-exo glycal 4b and 4c: Electrophilic

halogenation of 4a
Access to halo-exo-glycals 4b,c requires just one synthetic operation starting
from 4a. Accordingly, treatment of the latter with bromine/triethylamine or
iodonium dicollidinium triflate (IDCT)23 furnished halo-exo-glycals 4b and
4c, in 63% and 70% yields, respectively. The corresponding (Z)-derivatives
were obtained with complete stereoselectivity (Scheme 4a and 4b).
On the other hand, reaction of 4a with tetra-n-butylammonium tri-
bromide (nBu4NBr3, TBAT)24 furnished a 7:1 mixture of isomeric (Z)- and
(E)-4b (Scheme 4c). The latter mixture could be efficiently transformed into
dibromo exo-glycal 12, by further treatment with TBAT (Scheme 4d).The
stereochemical outcome of these reactions has already been rationalized25,26
Carbohydr. Chem., 2012, 38, 376–397 | 379

O Br2 O
Br
O NEt3 O
O O
CH2Cl2
(a)
63%
O O
4a 4b
O
I
O
IDCT O
(b) 4a
CH2Cl2
70% O
4c
O O
nBu4NBr3 Br
NEt3 O O
(c) 4a O + O Br
CH2Cl2
82% O O
4b (7 : 1) E-4b
nBu4NBr3 O
Br
NEt3 O
(d) 4b + E-4b O
Br
CH2Cl2
70% O
12
Scheme 4 Electrophilic halogenations of exo-glycals 4a and 4b.
and involves proton abstraction at C-1’ followed by base-induced, HBr or

HI, elimination from a preferred rotamer, in which the bulky halogen is
away from the oxirane ring.
3 Reactivity of functionalized furanose-based precursor 4a

As illustrated in Scheme 1, epoxy-exo-glycal 4a is a polyfunctionalized
derivative of potential rich reactivity. In this context, we set up to study its
chemistry by first evaluating the chemical behavior of the exocyclic olefin
and the oxirane ring as isolated functionalities, to subsequently study the
behavior of the vinyl oxirane moiety as a single unit.
3.1 Reactions at the exocyclic olefin

As part of the planned synthesis of 4b,c and within the context of evaluating
the reactivity of the enol ether-type olefin in 4a, we previously mentioned
the electrophilic halogenation of 4a.
3.1.1 Attempted hydrogenation of the enol ether double bond. The
hydrogenation of the exocyclic olefin was also studied since it could grant
access to synthetically valuable, functionalized, C-glycosyl compounds.27
Early attempts at hydrogenation of 4a (H2, Pd/C, EtOH, 25 psi) led to the
formation of furan derivative 13 (Scheme 5a) under a variety of hydro-
genation conditions.28 These results were in line with previous results from
our laboratory that already pointed to the easiness with which aromatization
380 | Carbohydr. Chem., 2012, 38, 376–397

O
O
O H2, Pd/C
O O
O
(a) EtOH
O 98%
4a 13
O O
O 1
O O NOE
O H2, Pd/C
2 H
EtOH
(b) HO HN
HO HN 95%
14 15
Scheme 5 Hydrogenation of exo-glycals 4a, and 14.
O
OH
1) BH3 . SMe2, THF
O
O
(a) 4a 2) NaOH, H2O2
81%
O
O
16
I
IDCT, H2O O
(b) 4a O
CH2Cl2
OH
37%
O
17
O
1) O3, CH2Cl2 O
O O
(c) 4a
2) SMe3
79%
O
18
Scheme 6 Several transformations of the exocyclic olefin in 4a.
takes place in these systems.29 In our opinion the aromatization process is

favored by the allylic oxirane moiety. The absence of the oxirane ring in these
molecules, e.g. 14, render the hydrogenation process possible (Scheme 5b).
The stereochemistry at the newly created stereogenic center in compound 15
was assigned by an observed NOE between H-2 and the methyl group at C-1.
3.1.2 Miscellaneous transformations at the exocyclic double bond. The
independent reactivity of the enol ether double bond was further demon-
strated by the transformations outlined in Scheme 6.19a Thus, hydrobora-
tion/oxidation of 4a, proved to be completely stereoselective to furnish C-1
glycoside 16, in 81% yield (Scheme 6a). Ketose 17 could also be accessed,
although in moderate yield, from 4a by treatment with IDCT in the pre-
sence of H2O (37%, Scheme 6b).30 Finally, ozonation of 4a permitted access
to synthetically useful epoxy-lactone 18, in 79% yield (Scheme 6c).
Carbohydr. Chem., 2012, 38, 376–397 | 381

3.2 Reactions at the oxirane ring
Nucleophilic ring opening reactions of furanose derived 2,3-oxiranes are
well documented. Lowary and co-workers31 have studied the regiochemistry
of the nucleophilic ring opening of isomeric alkyl 2,3-epoxy furanosides and
Hirota and co-workers32 studied the nucleophilic ring opening of fur-
anosidic allylic oxirane systems, more related to derivatives 4.
3.2.1 Reaction with oxygen nucleophiles. First, we examined the
nucleophilic ring opening of allylic oxiranes 4a–c with hydroxide anion. The
reaction took place smoothly upon treatment with nBu4NOH in refluxing
THF/H2O (3:1) to yield the corresponding diols 19a–c. (or diacetates
therefrom, 19b and 19c)19a in moderate yield with complete regio- and
stereoselectivity (Scheme 7).
3.2.2 Reaction with amines. The behavior of oxiranes 4 towards amines
proved to be similar, and their reactions with a variety of primary and
secondary amines in refluxing ethanol provided, in a completely regiose-
lective manner, 2-deoxy-2-amino derivatives in good yields. Displayed in
Scheme 8 are examples of the reaction of allylic oxiranes 4a–c with
morpholine (20) and 1-(2-methoxyphenyl)piperazine (21) to give allylic
amines 22a–c and 23a–c. In general the reactions were completed in 3–5 h.
All the reactions studied took place with complete stereoselectivity by
nucleophilic attack at the allylic C-2 position. Deshalo-epoxy-exo-glycal 4a
gave better yields of allylic amines than bromo- and iodo-exo-glycals 4b and
4c, respectively, with the latter giving the lower yields (e.g 22c and 23c
compared with 22b and 23b, Scheme 8).
Compounds 22 and 23 possess allylic amine substructures that have been
found in a variety of biologically active compounds.33
3.3 Pd(0)-Catalyzed reactions at the allylic oxirane moiety

The vinyl oxirane moiety in 4a made this compound an especially attractive
synthon for testing its reactivity in Pd(0)-catalyzed reactions by way of
its potential p-allyl palladium intermediate 24.34,35 With that in mind,
we studied the reaction of 4a with Pd(0) in the presence of nucleophiles
(4a-A),19b electrophiles (4a-B),36 and vinyl stannanes (4a-C)37 as
shown in Scheme 9.
3.3.1 Palladium catalyzed reaction of 4a with nucleophiles. p-Allylpal-
ladium intermediates, e.g., 24, are known to react with soft carbon
O O
X X
O i) nBu4NOH O
O O
THF/H2O (3:1), reflux
O RO OR
ii) Ac2O, py for 19b and 19c
4a 19a (R = H, 57%)
4b 19b (R = Ac, 59%)
4c Series: a X = H; b X = Br; c X = I 19c (R = Ac, 64%)
Scheme 7 Reaction of allylic oxiranes 4a–c with nBu4NOH leading to diols, or diacetates
therefrom.
382 | Carbohydr. Chem., 2012, 38, 376–397

O
O
O X
X O
O N O
20 H
O
(a)
EtOH, reflux HO N
O
O
4a 22a (85%)
4b 22b (79%)
4c 22c (72%)
HN O
X
OMe O
O N O
X
O
O
HO N
(b) 21
O EtOH, reflux N OMe
4a 23a (88%)
4b 23b (80%)
4c Series: a X = H; b X = Br; c X = I 23c (63%)
Scheme 8 Reaction of epoxy-exo-glycals 4a–c with amines in refluxing ethanol.
O
Nu
O
O
NuH
HO
A
O O
E
O O
Pd(0) O ZnEt2 O
4a Pd Electrophile
L L (E) HO
O B R1
24
O
R3Sn
O
O
R1
HO
C
Scheme 9 Proposed Pd(0)-catalyzed reactions of vinyl oxirane 4a with nucleophiles, electro-

philes, and vinylstannanes.
nucleophiles and amines, to give substituted allylic alcohols, e.g., A

(Scheme 9). In order to test this transformation, we first treated vinyloxirane
4a with Pd(OAc)2 and PPh3 in THF at rt, overnight, in the presence of
methyl malonate. Compound 25, which had resulted from 1,4 addition
to the vinyloxirane system could then be isolated.19b Likewise, treatment
of diverse primary and secondary amines yielded allylic derivatives in
moderate to good yields, e.g. 26, 27 (Scheme 10).
Carbohydr. Chem., 2012, 38, 376–397 | 383

O
CO2Me
MeO2C O
O CO2Me
Pd(OAc)2, PPh3, THF, rt

CO2Me
86% HO 25
O O
O 20 O
O O
N
O
O 81% HO 26
4a O
21 O
O
N
86% HO 27 N
MeO
Scheme 10 Pd(0)-catalyzed reactions of vinyl oxirane 4a with methyl malonate, morpholine

(20), and 1-(2-methoxyphenyl)piperazine (21).
O O R2
R1 O OH
R1
O O O O
O Et2Zn O O
Pd Zn R2
O L L O
HO
24 28 29
Scheme 11 Proposed reaction pathway for Pd(0)/Et2Zn-catalyzed reaction between p-allyl-

palladium species 24 and electrophilic carbonyl compounds.
3.3.2 Palladium catalyzed reaction of 4a with electrophiles. In general,

p-allylpalladium complexes are of electrophilic nature, however, several
examples of polarity inversion in p-allylpalladium species by transmetalla-
tion with low valence metals have appeared.38 More recently, Et3B and
Et2Zn have been used in the Pd-catalyzed generation of nucleophilic allyl
species from allyl esters and ethers.39
In our case, we found that allylzinc derivatives, e.g. 28, readily obtained
from p-allylpalladium species 24 by ligand transfer to Et2Zn, were capable
of reacting with aldehydes and ketones to give functionalized C-1 glycals 29
(Scheme 11).40
From a experimental standpoint the reaction was carried out as follows, a
solution containing vinyl epoxide 4a, Pd(PPh3)4, and the corresponding
carbonyl compound, in THF, at rt, was treated with Et2Zn. These reactions
take place smoothly, and dissapearance of the starting material normally
occurred within 1–3 h.
This Pd(0)-catalyzed transformation works well with aldehydes, e.g. 30,
31, and moderately well with ketones, e.g. 32, 33, and with excellent 1,4-
regioselectivity (Scheme 12). o-Methoxycinnamaldehyde (30) yielded a 1.5:1
diastereomeric mixture of allylic alcohols 29a (Scheme 12). A slightly higher
384 | Carbohydr. Chem., 2012, 38, 376–397

OHC
MeO O
(1.5:1)
O
O
30
HO
HO MeO
Pd(PPh3)4, Et2Zn, THF
29a
70%
CHO
O
(2.5:1)
O
O
31
O HO
HO
O 29b
O 72%
O
O
O O
32 O HO
4a HO
29c
48%
O
O
O
O Ph
Ph Ph
33
HO Ph
Pd(PPh3)4, Et2Zn, THF HO
29d
68%
Scheme 12 Reaction of oxirane 4a with aldehydes and ketones mediated by Pd(0)/Et2Zn in

THF.
diastereomeric ratio (2.5:1) was observed in the reaction of 4a with citro-

nellal (31). Ketones 32 and 33, gave moderate yields of tertiary hydroxy
derivatives 29c and 29d.
3.3.3 Palladium catalyzed reaction of 4a with vinyl stannanes. Based on
the significant work of Stille’s,41 and Echavarren’s42 groups on the use of
stannanes as mild nucleophiles in palladium catalyzed couplings with vinyl
epoxides we decided to explore the reactivity of compound 4a.
Accordingly, the reaction of 4a with commercially available tetravinyl tin
under the conditions recommended by Stille and co–workers (Scheme 13)41b
took place smoothly and furnished a mixture of regioisomeric derivatives 34
and 35, resulting from 1,4- and 1,2-addition to the allylic system, respec-
tively. In keeping with literature precedents, compound 35, generated by
1,2-addition was obtained as the minor isomer (1.4:1 ratio).37 The structural
assignments were based on the study of their 1H- and 13C NMR spectra,
and the configuration at C-2 in compound 35, already inferred by
mechanistic reasons, was confirmed by a NOESY experiment that showed
the correlation of H-3 with the non-terminal vinylic proton.
The flexibility of this reaction was tested with more complex vinyl stan-
nanes, e.g. 36–38, readily available by Et3B catalyzed radical addition of
Bu3SnH to terminal alkynes (Fig. 3).43
Carbohydr. Chem., 2012, 38, 376–397 | 385

O
O
(CH2=CH)4Sn 1’
O
O 1’ O
Pd(CH3CN)2Cl2 O
4a 3 + 2
H2O (10 eq.)
HO H3
DMF HO
observed NOESY H
34 (51%) 35 (15%)
Scheme 13 Reaction of vinyloxirane 4a, with tetravinyl tin catalyzed by Pd(CH3CN)2Cl2.
O
O O N N
Bu3Sn
SnBu3
Bu3Sn O O
36 37 38
Fig. 3 Vinyl stannanes 36–38, prepared by Et3B catalyzed radical addition of Bu3SnH to
alkynes.
O O
O
O O
O O O
O O
HO O HO
O
O O
O O O
O
(From 4a + 36) 39 (4.6 : 1) 40 67% yield
O
O
O
O
O
O
HO H
N 3
HO H N
NOESY
(From 4a + 37) 41 (3 : 1) 42 81% yield
O
O
O N
O HO H3 N
O
NOESY H
HO
(From 4a + 38) 43 (2.5 : 1) 44 70% yield
Fig. 4 Reaction products of vinyl oxirane 4a with vinyl stannanes 36–38.
According to that, alkenyl stannanes 36–38 (Fig. 3) react with 4a under

the above mentioned Stille conditions to give the compounds displayed in
Fig. 4. In all cases, a mixture of 1,4- and 1,2-regioisomers was obtained.
The C-1 glycal derivatives (1,4-isomers) were always obtained as the major
386 | Carbohydr. Chem., 2012, 38, 376–397

isomers and proved to be very acid sensitive, then leading to the corre-
sponding furan derivatives, however the use of Et3N (3%) in the chroma-
tography purifications helped to solve that issue.
4 Reactivity of the alkenyl halide moiety in functionalized furanose-based

precursors 4b,c
In addition to the functionality present in exo-glycal 4a, compounds 4b or
4c possess an alkenyl halide which could function as an additional handle
for incorporating molecular diversity at C-1’. In this context we carried out
preliminary studies on Pd-catalyzed reactions of structurally related
furanose-derived halo-exo-glycals 19 (devoid of the oxirane moiety).
4.1 Cross-coupling reactions of 1’-halo-exo-glycals

We had previously identified the Suzuki-Miyaura44 and the Sonoga-
shira45,46 cross-coupling reactions as synthetically useful transformations,
and we had shown that they could be performed in furanose 1’-halo-exo-
glycals, e.g. 19b,c, leading to substituted exo-glycals, e.g. 45, 46, respec-
tively47,48 (Scheme 14).
These reactions took place with retention of the configuration at the
halide, and led to exo-glycals in moderate to good yields. Alkenyl iodides
led to higher yields of exo-glycals than alkenyl bromides in both Suzuki-
Miyaura (Scheme 14a,b) and Sonogashira couplings (Scheme 14c,d).
4.2 Attempted palladium-catalyzed allylic amination of 1’halo-alkenyl

derivatives 4b
Furanose exo-glycal 4a, had revealed itself as a valuable substrate, upon
treatment with Pd(0), in the preparation of a variety of highly functionalized
C-1 glycals.
Scheme 14 Suzuki-Miyaura and Sonogashira cross-coupling reactions of 1’halo-exo-glycals

19b,c.
Carbohydr. Chem., 2012, 38, 376–397 | 387

Scheme 15 Reaction of compounds 4a and 4b with amines.
In order to compare its behavior, i.e. 4a-47 (Scheme 15), in palladium

catalyzed aminations versus that of 1’halo derivative 4b, we carried out the
reaction of 4b with piperidine, under experimental conditions similar to those
successfully applied to 4a [Pd(OAc)2, PPh3, THF, rt (Scheme 15)]. Under
these reaction conditions, formation of the expected product 48 was not
observed and the starting material was recovered unreacted. It was found,
however, that refluxing of the above mentioned solution provided amino-
alcohol 49, in 80% yield (Scheme 15). The fact that the stereochemistry at
C-1’ was maintained in allylic amine 49, suggested that no p-allyl palladium
intermediate had been involved in the transformation. This was indeed
confirmed when a refluxing ethanolic solution containing piperidine and 4b
yielded exclusively allylic amine 49, in 88% yield (Scheme 15).
These observations seemed to indicate that the reaction taking place was
an uncatalyzed nucleophilic opening of the epoxide, and that the presence of
bromine in the vinyl oxirane deactivates the system towards allylic ionization
with palladium.49–51
5 Three components assembly of functionalized furanose derivatives 4b,c

From the above mentioned studies, we were aware that: i) uncatalyzed
amination could be carried out in vinyl oxiranes 4a 4b and 4c, leading to 2-
deoxy-2-amino furanose derivatives, e.g. 4a–c-22a–c (Scheme 8), and ii)
that, whereas palladium-catalyzed amination could be performed in deshalo-
exo-glycal 4a, it was not effective in alkenyl bromide 4b (Scheme 15).
388 | Carbohydr. Chem., 2012, 38, 376–397

Scheme 16 Potential three-components assemblies of 1’halo-epoxy-exo glycal 4b.
With these data in hand, we envisaged that the transformations outlined

in Scheme 16 could be feasible, since uncatalyzed nucleophilic opening of
the oxirane ring could be followed by cross-coupling reaction of the alkenyl
halide, or vice versa, cross-coupling on the alkenyl bromide could be fol-
lowed by (uncatalyzed) amination.
5.1 One-pot multicomponent amination/Suzuki cross-coupling of 4b

We found that the treatment of 4b with Pd(PPh3)4, NaOH, in THF:H2O, at
60 1C, allowed its multicomponent assembly with boronic acids and amines to
give access to a library of compounds with three sites for molecular diversity.
In all instances where a comparison was possible, the one-pot transfor-
mation was more efficient than the alternative stepwise synthesis. This
transformation could be applied to a variety of aryl and heteroaryl boronic
acids (Table 1).
Alkyl boranes have also been used in B-alkyl Suzuki cross-coupling
reactions.52 So, the methodology could also be applied to the coupling of 4b
with alkyl boranes and secondary amines, as shown in Table 2. The reaction
of vinyl bromide 4b with diethyl(3-pyridyl)borane and different cyclic
amines led to derivatives 55–57, in good yields.
Unlike the success achieved with the three component reactions involving
Suzuki cross-couplings (Tables 1 and 2), the Sonogashira variant did not
lead to the desired products. For instance, instead of the desired amino
alkynes 50b (Scheme 16), the reaction of compound 4b with phenyl acet-
ylene, or trimethylsilylacetylene [alkyne (1.1 eq.), Pd(PPh3)4 (5%), CuI
(10%), Et2NH (4 mL/mmol), rt] produced a mixture of 2-deoxy enynes 58
[1.2:1, unassigned (Z:E) mixture] and 59 [2:1, unassigned (Z:E) mixture],
respectively (Table 3, entries i, ii). Likewise, when dodec-1-yne was
employed, a E1:1 mixture of enynes 60 (37%) was obtained, although this
time accompanied by 2-deoxy-2-amino derivative 61 (17%) (Table 3, entry
iii). These results seemed to indicate that the Sonogashira cross-coupling
was taking place first and it was followed by the formation of a p-allyl
palladium complex responsible for the formation of compounds 58–60,
as geometric enyne mixtures.w Furthermore, the addition of hydride
w
Will no p-allyl palladium complex be involved, compound 30 would have appeared as a single
isomer, owing to the retention of configuration associated with the Sonogashira cross-coupling
reaction of alkenyl halides.
Carbohydr. Chem., 2012, 38, 376–397 | 389

Table 1 One-pot multicomponent assembly of 1’-alkenyl exo-glycal 4b with piperidine and
boronic acids leading to a library of functionalized allylic amines.
entry R3-B(OH)2 product yield (%)
i 51 63
ii 52 60
iii 53 72
iv 54 75
Table 2 One-pot multicomponent assembly of furanose derivatives 55–57, by reaction of

alkenyl bromide 4a, with diethyl(3-pyridyl)borane and amines.
entry R1R2NH product yield (%)
i 55 78
ii 56 71
iii 57 74
from a palladium-hydride species to the p-allyl palladium could be

responsible for the formation of the 2-deoxy, rather than 2-deoxy-2-amino
derivatives.53
390 | Carbohydr. Chem., 2012, 38, 376–397

Table 3 Sonogashira cross-coupling reaction of epoxy bromo-exo-glycal 4b, with phenyl
acetylene, trimethylsilylacetylene, and dodec-1-yne, in the presence of amines.
Entry Reaction conditions Products (yield)
ii
iii
iv
5.2 Attempted one-pot multicomponent amination/Sonogashira

cross-coupling48b
The presence of an unreacted bromine atom in amino alcohol 61 (Table 3,
entry iii) was unexpected. Therefore, we carried out several attemps to
conduct Sonogashira couplings with 61 and phenylacetylene that were not
successful. When the reaction was carried out in the presence of piperidine
at 60 1C (Table 3, entry iv), only amino alcohol 62 could be isolated.
6 Generation of furanosidic libraries possessing three – or more – sites for

molecular diversity
The furanose derivatives 51–54 and 55–57, (Tables 1 and 2, respectively)
resulting from oxirane opening followed by palladium catalyzed Suzuki
cross-coupling, incorporate already two potential pharmacophores. In
addition, the presence of the 3-OH group and the 5,6-O-isopropylidene
acetal makes possible further derivatization of these derivatives.
In order to illustrate this point, amino alcohols 55 and 56 can be trans-
formed into 63, 64 and 65 (Scheme 17). Thus, laying the bases for the
synthesis of additional furanose-based libraries.
Carbohydr. Chem., 2012, 38, 376–397 | 391

N N
O HO
O O
O HO
AcOH/THF/H2O
75%
RO N O N
O
64
Ac2O/py 55 R = H
63 R = Ac
97%
N N
O HO
O O
O AcOH/THF/H2O HO
56%
HO N HO N
56 O 65 O
Scheme 17 Furanose derivatives 64 and 65, with more than three points of diversity.
Acetylation of 55 led to acetate 63 with three points of diversity. Inter-

estingly, the enol ether double bond in 63 displayed an enhanced stability
towards acid, and consequently the 5,6-O-isopropylidene acetal group could
be removed without tampering with the anomeric functionality, leading to
diol 64.
These compounds already possess more than three sites for structural
diversity, and the 5,6-diol could serve as a handle for subsequent incor-
poration of molecular diversity.
Likewise, treatment of amino alcohol 56 with AcOH:THF:H2O (4:2:1)
under reflux, yielded triol 65.
7 Conclusions
Highly functionalized exo-glycal 4a is a useful synthetic intermediate in the
generation of molecular diversity in a chemoselective manner. In this report,
we have shown that it could be successfully used in the preparation of
furanose-based libraries with three or more sites for molecular diversity.
Compound 4a contains an allylic epoxide moiety that can undergo che-
moselective reactions either at the double bond, at the oxirane ring, or at the
vinyl epoxide system.
Halogenation of the double bond gives rise to synthetically useful 1’-halo-
epoxy-exo-glycals (4b,c). Hydrogenation of 4a led to the rapid formation of
furan derivatives, which could be circumvented by performing the hydro-
genation on exo-glycals devoid of the oxirane ring. The hydrogenation,
then, takes place in a stereocontrolled manner, probably due to the presence
of the bulky b-oriented 5,6-O-isopropylidene ring.
The oxirane moiety undergoes ring opening with nitrogen nucleophiles or
hydroxide in a completely regioselective manner, leading to allylic amino
alcohols and allylic diols, respectively.
392 | Carbohydr. Chem., 2012, 38, 376–397

Nucleophilic ring opening of the oxirane ring in 1’-halo-epoxy-exo-gly-
cals leaves the vinyl halide functionality unreacted, so that it can be engaged
in Suzuki cross-coupling reactions with boronic acids, thus allowing the
generation of a library of furanose-based allylic amines. These amines still
bear a OH-3, and a 5,6-isopropylidene acetal ready for further transfor-
mations. Acylation can then be performed at O-3, to incorporate a third
substituent. Remarkably, these derivatives undergo acid-mediated chemo-
selective 5,6-O-isopropylidene ring opening without aromatization to the
corresponding furan derivatives.21 Thus diols, e.g. 64, and triols, e.g. 65,
ready for further derivatization can be readily prepared.
The rich chemistry of these derivatives still merits further scrutiny, and in
our opinion further transformation can still be uncovered allowing access to
templates with up to five sites of molecular diversity.
Abbreviations
DMF Dimethylformamide
h Hour
IDCT Iodonium dicollidinium triflate
NOE Nuclear overhauser effect
rt Room temperature
TBAT Tetra-n-butylammonium tribromide
THF Tetrahydrofuran
Acknowledgements
This research has been supported with funds from the Ministerio de Ciencia
e Innovación grants CTQ-2006-C03 and CTQ2009-10343, and Comunidad
de Madrid grant S2009/PPQ-1752. Generous financial support from
Janssen-Cilag S.A. is also acknowledged. Additional support came as
predoctoral scholarships to Ana Pedregosa (Janssen-Cilag S.A) and
Aitor Barrio [Consejo Superior de Investigaciones Cientı´ficas (CSIC)] who
were intensely involved in the development of this research.
References
1 (a) W. Meutermans, G. T. Le, and B. Becker, Chem. Med. Chem., 2006, 1,
1164. (b) G. T. Le, G. Abbenante, B. Becker, M. Grathwohl, J. Halliday, G.
Tometzki, J. Zuegg and W. Meutermans, Drug Discovery Today, 2003, 8, 701.
2 (a) L. Cipolla and F. Peri, Mini-Rev. Med. Chem., 2011, 11, 39. (b) L. Cipolla,
A. C. Araujo, D. Bini, L. Gabrielli, L. Russo and N. Shaikh, Expert Opin. Drug
Discov., 2010, 5, 723. (c) G. Abbenante, B. Becker, S. Blanc, C. Clark, G.
Condie, G. Fraser, M. Grathwohl, J. Halliday, S. Henderson, A. Lam, L. Liu,
M. Mann, C. Muldoon, A. Pearson, R. Premraj, T. Ramsdale, T. Rossetti, K.
Schafer, G. L. Thanh, G. Tometzki, F. Vari, G. Verquin, J. Waanders, M.
West, N. Wimmer, A. Yau, J. Zuegg and W. Meutermans, J. Med. Chem., 2010,
53, 5576. (d) L. Cipolla, F. Peri, B. La Ferla, C. Redaelli and F. Nicotra, Curr.
Org. Synth., 2005, 2, 153.
3 B. Becker, G. C. Condie, G. T. Le and W. Meutermans, Mini-Rev. Med. Chem.,
2006, 6, 1299.
Carbohydr. Chem., 2012, 38, 376–397 | 393

4 (a) R. Hirschmann, K. C. Nicolaou, S. Pietranico, J. Salvino, E. M. Leahy, P.
A. Sprengeler, G. Furst, A. B. Smith III, C. D. Strader, M. A. Cascieri, M. R.
Candelore, C. Donaldson, W. Vale and L. Maechler, J. Am. Chem. Soc., 1992,
114, 9217. (b) R. Hirschmann, K. C. Nicolaou, S. Pietranico, E. M. Leahy, J.
Salvino, B. Arison, M. A. Cichy, P. G. Spoors, W. C. Shakespeare, P. A.
Sprengeler, P. Hamley, A. B. Smith III T. Reisine, K. Raynor, L. Maechler, C.
Donaldson, W. Vale, R. M. Freidinger, M. A. Cascieri and C. D. De Strader,
J. Am. Chem. Soc., 1993, 115, 12550. (c) R. Hirschmann, W. Q. Yao, M. A.
Cascieri, C. D. Strader, L. Maechler, M. A. Cichy-Knight, J. Hynes, Jr, R. D.
van Rijn, P. A. Sprengeler and A. B. Smith III, J. Med. Chem., 1996, 39, 2441.
(d) R. F. Hirschmann, K. C. Nicolaou, A. R. Angeles, J. S. Chen, and A. B.
Smith III, Acc. Chem. Res., 2009, 42, 1511.
5 J. Boer, D. Gottschling, A. Schuster, B. Holzmann and H. Kessler, Angew.
Chem. Int. Ed., 2001, 40, 3870.
6 S. Hanessian, N. Moitessier and S. Wilmouth, Tetrahedron, 2000, 56, 7643.
7 For reviews on carbohydrate-derived structures as templates, see: (a) Y. M.
Chabre and R. Roy, in: The Sugar Code. Fundamentals of Glycosciences (Ed.:
H.-J. Gabius), Wiley-VCH, Weinheim, Germany, 2009, p. 53. (b) P. V.
Murphy and T. Velasco-Torrijos, in: Glycoscience – Chemistry and Chemical
Biology, 2nd ed. (Eds.: B. Fraser-Reid, K. Tatsuta and J. Thiem), Springer-
Verlag, Berlin-Heidelberg, 2008; p. 997. (c) P. V. Murphy, Eur. J. Org. Chem.,
2007, 4177. (d) I. Velter, B. La Ferla and F. Nicotra, J. Carbohydr. Chem.,
2006, 25, 97. (e) P. V. Murphy and J. L. Dunne, Curr. Org. Synth., 2006, 3, 403.
(f) T. Kanemitsu and O. Kanie, Trends Glycosci. Glycotechnol., 1999, 11, 267.
(g) M. J. Sofia and D. J. Silva, Curr. Opin. Drug Discovery Dev., 1999, 2, 365.
8 Pyranose-based scaffolds: (a) N. Elmouelhi, U. Aich, V. D. P. Paruchuri, M. A.
Meledeo, C. T. Campbell, J. J. Wang, R. Srinivas, H. S. Khanna and K. J.
Yarema, J. Med. Chem., 2009, 52, 2515. (b) C. T. Campbell, U. Aich, C. A.
Weier, J. J. Wang, S. S. Choi, M. M. Wen, K. Maisel, S.-G. Sampathkumar and
K. J. Yarema, J. Med. Chem., 2008, 51, 8135. (c) Y. Tamaruya, M. Suzuki, G.
Kamura, M. Kanai, K. Hama, K. Shimizu, J. Aoki, H. Arai and M. Shibasaki,
Angew. Chem. Int. Ed., 2004, 43, 2834. (d) N. Moitessier, S. Dufour, F.
Chretien, J. P. Thiery, B. Maigret and Y. Chapleur, Bioorg. Med. Chem., 2001,
9, 511. (e) P. V. Murphy, J. L. O’Brien, L. J. Gorey-Feret and A. B. Smith III,
Bioorg. Med. Chem. Lett., 2002, 12, 1763. (f) L. Cipolla, C. Redaelli, F.
Granucci, G. Zampella, A. Zaza, R. Chisci and F. Nicotra, Carbohydr. Res.,
2010, 345, 1291. (g) C. Rosenbohm, D. Van den Berghe, A. Vlietinck and
J. Wengel, Tetrahedron, 2001, 57, 6277.
9 Furanose-based scaffolds: (a) C. Papageorgiou, R. Haltiner, C. Bruns and
T. J. Petcher, Bioorg. Med. Chem. Lett., 1992, 2, 135. (b) T. K. Chakraborty,
S. Ghosh, M. H. V. R. Rao, A. C. Kunwar, H. Chob and A. K. Ghosh,
Tetrahedron Lett., 2000, 41, 10121. (c) E. B. Krueger, T. P. Hopkins, M. T.
Keaney, M. A. Walters and A. M. Boldi, J. Comb. Chem., 2002, 4, 229. (e) A.
Novoa, N. Pellegrini-Moı̈se, D. Bechet, M. Barberi-Heyob and Y. Chapleur,
Bioorg. Med. Chem., 2010, 18, 3285. (f) A. A. Edwards, O. Ichihara, S. Murfin,
R. Wilkes, M. Whittaker, D. J. Watkin and G. W. J. Fleet, J. Comb. Chem.,
2004, 6, 230.
10 S. Castoldi, M. Cravini, F. Micheli, E. Piga, G. Russo, P. Seneci and L. Lay,
Eur. J. Org. Chem., 2004, 2853.
11 Sugar-based bicyclic scaffolds: (a) F. Peri, R. Bassetti, E. Caneva, L. De Gioia,
B. La Ferla, M. Presta, E. Tanghetti and F. Nicotra, J. Chem. Soc., Perkin
Trans. I, 2002, 5, 638. (b) A. C. Araujo, F. Nicotra, B. Costa, G. Giagnoni and
L. Cipolla, Carbohydr. Res., 2008, 343, 1840. (c) A. C. Araujo, F. Nicotra, C.
394 | Carbohydr. Chem., 2012, 38, 376–397

Airoldi, B. Costa, G. Giagnoni, P. Fumagalli and L. Cipolla, Eur. J. Org.
Chem., 2008, 635. (d) G. Cervi, F. Peri, C. Battistini, C. Gennari and F.
Nicotra, Bioorg. Med. Chem., 2006, 14, 3349. (e) A. Cordeiro, E. Quesada,
M.-C. Bonache, S. Velazquez, M.-J. Camarasa and A. San-Felix, J. Org. Chem.,
2006, 71, 7224. (f) G. Capozzi, S. Giannini, S. Menichetti, C. Nativi, A. Gio-
litti, R. Patacchini, E. Perrotta, M. Altamurac and C. A. Maggic, Bioorg. Med.
Chem. Lett., 2002, 12, 2263. (g) J. Jimenez-Barbero, E. Dragoni, C. Venturi,
F. Nannucci, A. Arda, M. Fontanella, S. Andre, F. J. Cañada, H.-J. Gabius
and C. Nativi, Chem. Eur. J., 2009, 15, 10423. (h) S. Mari, F. J. Cañada,
J. Jimenez-Barbero, A. Bernardi, G. Marcou, I. Motto, I. Velter, F. Nicotra and
B. La Ferla, Eur. J. Org. Chem., 2006, 2925.
12 (a) H. Hünger, J. Ohnsmann and H. Kunz, Angew. Chem. Int. Ed., 2004, 43,
1104. (b) T. Opatz, C. Kallus, T. Wunberg, W. Schmidt, S. Henke and H.
Kunz, Eur. J. Org. Chem., 2003, 1527. (c) T. Wunberg, C. Kallus, T. Opatz, S.
Henke, W. Schmidt and H. Kunz, Angew. Chem. Int. Ed., 1998, 37, 2503.
13 (a) R. Jain, M. Kamau, C. Wang, R. Ippolito, H. Wang, R. Dulina,
J. Anderson, D. Gange and M. J. Sofia, Bioorg. Med. Chem. Lett., 2003, 13,
2185. (b) M. J. Sofia, R. Hunter, T. Y. Chan, A. Vaughan, R. Dulina, H. Wang
and D. Gange, J. Org. Chem., 1998, 63, 2802.
14 N. Moitessier, C. Henry, N. Aubert and Y. Chapleur, Tetrahedron Lett., 2005,
46, 6191.
15 (a) A. M. Gómez, A. Pedregosa, M. Casillas, C. Uriel and J. C. López, Eur. J.
Org. Chem., 2009, 3579. (b) A. M. Gómez, M. Casillas, S. Valverde and J. C.
López, J. Chem. Soc., Chem. Commun., 1996, 2357.
16 A. Basha, R. Henry, M. A. McLaughlin, J. D. Ratajczyk and S. J. Witten-
berger, J. Org. Chem., 1994, 59, 6103.
17 J. Gelas and D. Horton, Carbohydr. Res., 1978, 67, 371.
18 R. E. Ireland, S. Thaisrivongs, N. Vanier and C. S. Wilcox, J. Org. Chem., 1980,
45, 48.
19 (a) A. M. Gómez, A. Pedregosa, C. Uriel, S. Valverde and J. C. López, Eur. J.
Org. Chem., 2010, 5619. (b) A. M. Gómez, A. Pedregosa, S. Valverde and J. C.
López, Chem. Commun., 2002, 2022.
20 R. U. Lemieux and B. Fraser-Reid, Can. J. Chem., 1965, 43, 1460.
21 (a) P. Crotti, V. Di Busolo, L. Favero, F. Macchia and M. Pineschi, Eur. J.
Org. Chem., 1998, 1675. (b) P. Herdewijn, K. Ruf and W. Pfleiderer, Helv.
Chim. Acta 1991, 74, 7. (c) B. Fraser-Reid, L. Magdzinski, B. Molino and D.
R. Mootoo, J. Org. Chem., 1987, 52, 4495. (d) B. Fraser-Reid, L. Magdzinski
and B. Molino, J. Am. Chem. Soc., 1984, 106, 731.
22 J.-P. Praly, G.-R. Chen, J. Gola and G. Hetzer, Eur. J. Org. Chem., 2000, 2831.
23 (a) G. H. Veeneman, S. H. van Leeuwen, H. Zuurmond and J. H. van Boom,
J. Carbohydr. Chem., 1990, 9, 783. (b) D. Noort, G. H. Veeneman, G.-J. Boons,
G. A. van der Marel, G. J. Mulder and J. H. van Boom, Synlett, 1990, 205.
24 G. Bellucci, C. Chiappe and G. Lo Moro, J. Org. Chem., 1997, 62, 3176.
25 (a) M. Lakhrissi, G. Carchon, T. Schlama, C. Mioskowski and Y. Chapleur
Tetrahedron Lett., 1998, 39, 6453. (b) W.-B. Yang, Y.-Y. Yang, Y.-F. Gu, S.-H.
Wang, C.-C. Chang and C.-H. Lin, J. Org. Chem., 2002, 67, 3773.
26 A. M. Gómez, A. Pedregosa, S. Valverde and J. C. López, Tetrahedron Lett.,
2003, 44, 6111.
27 (a) X.-P. He, X.-W. Wang, X.-P. Jin, H. Zhou, X.-X. Shi, G.-R. Chen and
Y.-T. Long, J. Am. Chem. Soc., 2011, 133, 3649. (b) S. I. van Kasteren, H. B.
Kramer, H. H. Jensen, S. J. Campbell, J. Kirkpatrick, N. J. Oldham, D. C.
Anthony and B. G. Davis, Nature, 2007, 446, 1105. (c) M. A. Fischbach, H.
Lin, D. R. Liu and C. T. Walsh, Proc. Natl. Acad. Sci. U.S.A., 2005, 102, 571.
Carbohydr. Chem., 2012, 38, 376–397 | 395

28 A. M. Gómez, A. Pedregosa, C. Uriel, S. Valverde and J. C. López, Eur. J. Org.
Chem., 2010, 5619.
29 A. M. Gómez, M. Casillas, B. Rodrı́guez, S. Valverde and J. C. López,
ARKIVOC, 2010, iii, 288.
30 For reactions of exo-glycals with iodonium ions, see: a) L. Lay, F. Nicotra, L.
Panza and G. Russo, Synlett, 1995, 167. b) A. Haudrechy and P. Sinaÿ, Car-
bohydr. Res., 1991, 216, 375.
31 Ring opening of alkyl 2,3-epoxy furanosides, see: (a) D. Hou and T. L.
Lowary, Org. Lett., 2007, 9, 4487. (b) Y. Bai and T. L. Lowary, J. Org.
Chem., 2006, 71, 9658. (c) R. R. Gadikota, C. S. Callam, T. Wagner, B. Del
Fraino and T. L. Lowary, J. Am. Chem. Soc., 2003, 125, 4155. (d) C. S.
Callam, R. R. Gadikota, D. M. Krein and T. L. Lowary, J. Am. Chem. Soc.,
2003, 125, 13112. (e) C. S. Callam, R. R. Gadikota and T. L. Lowary,
Synlett, 2003, 1271.
32 Ring opening of furanose-derived allylic epoxides, see: (a) H. Sajiki, H. Takasu
and K. Hirota, Heterocycles 2006, 67, 399. (b) H. Takasu, Y. Tsuji, H. Sajiki
and K. Hirota, Tetrahedron, 2005, 61, 11027. (c) H. Takasu, Y. Tsuji, H. Sajiki
and K. Hirota, Tetrahedron, 2005, 61, 8499. (d) K. Hirota, H. Takasu, Y. Tsuji
and H. Sajiki, Chem. Commun., 1999, 1827.
33 V. Nesterenko, K. S. Putt and P. J. Hergenrother, J. Am. Chem. Soc., 2003, 125,
14672.
34 (a) J. Tsuji, H. Kataoka and Y. Kobayashi, Tetrahedron Lett., 1981, 22, 2575.
(b) B. M. Trost and G. A. Molander, J. Am. Chem. Soc., 1981, 103, 5969.
35 B. M. Trost and D. L. Van Kranken, Chem. Rev., 1996, 96, 395.
36 A. M. Gómez, A. Barrio, A. Pedregosa, S. Valverde and J. C. López, Tetra-
hedron Lett., 2003, 44, 8433.
37 A. M. Gómez, A. Pedregosa, A. Barrio, S. Valverde and J. C. López, Eur. J.
Org. Chem., 2009, 4627.
38 These reactions include treatment with chromium: O. Fujiwara, K. Takai and
K. Utimoto, J. Org. Chem., 1990, 55, 1705; tin: Y. Masuyama, J. Nakata and
Y. Kurusu, J. Chem. Soc., Perkin Trans. 1, 1991, 2598; samarium: J. M.
Aurrecoechea and E. Iztueta, Tetrahedron Lett., 1995, 36, 7129; and indium:
S. Araki, K. Kameda, J. Tanaka, T. Hirashita, H. Yamamura and M. Kawai,
J. Org. Chem., 2001, 66, 7919; S. Araki, T. Kamei, T. Hirashita, H. Yamamura
and M. Kawai, Org. Lett., 2000, 2, 847; T. Tabuchi, J. Inanaga and
M. Yamaguchi, Tetrahedron Lett., 1986, 27, 1195.
39 Review: Y. Tamaru, Eur. J. Org. Chem., 2005, 2647, and references cited
therein.
40 A. M. Gomez and J. C. Lopez, Specialist Periodical Reports, Carbohydrate
Chemistry, 2009, 35, 290. DOI:10.1039/B901578G.
41 (a) A. M. Echavarren, D. R. Tueting and J. K. Stille, J. Am. Chem. Soc., 1988,
110, 4039. (b) D. R. Tueting, A. M. Echavarren and J. K. Stille, Tetrahedron,
1989, 45, 979.
42 (a) A. M. Castaño, M. Ruano and A. M. Echavarren, Tetrahedron Lett., 1996,
37, 6591. (b) A. M. Castaño, M. Méndez, M. Ruano and A. M. Echavarren,
J. Org. Chem., 2001, 66, 589.
43 (a) K. Nozaki, K. Oshima and K. Utimoto, J. Am. Chem. Soc., 1987, 109,
2547. (b) K. Nozaki, K. Oshima and K. Utimoto, Tetrahedron, 1989, 45, 923.
44 (a) N. Miyaura and A. Suzuki, Chem. Rev., 1995, 95, 2457. (b) A. Suzuki,
J. Organomet. Chem., 1999, 576, 147.
45 K. Sonogashira, J. Organomet. Chem., 2002, 653, 46, and references cited
therein.
46 Review: R. Chinchilla and C. Nájera, Chem. Rev., 2007, 107, 874.
396 | Carbohydr. Chem., 2012, 38, 376–397

47 Suzuki-Miyaura cross-coupling: A. M. Gómez, G. O. Danelon, A. Pedregosa,
S. Valverde and J. C. López, Chem. Commun., 2002, 2024.
48 Sonogashira cross-coupling: (a) A. M. Gómez, A. Pedregosa, A. Barrio,
S. Valverde and J. C. López, Tetrahedron Lett., 2004, 45, 6307. (b) A. M.
Gómez, A. Barrio, A. Pedregosa, C. Uriel, S. Valverde and J. C. López, Eur. J.
Org. Chem., 2010, 2910.
49 For studies on allylic ionization versus oxidative addition into vinyl C-Br
bonds, see: (a) E. Comer, M. G. Organ and S. J. Hynes, J. Am. Chem. Soc.,
2004, 126, 16087. (b) M. G. Organ, E. A. Arvanitis, C. E. Dixon and J. T.
Cooper, J. Am. Chem. Soc., 2002, 124, 1288.
50 For studies on 1-acetoxy-2-bromo-2-alkenes: (a) G. C. Nwokogu, J. Org.
Chem., 1985, 50, 3900. (b) G. C. Nwokogu, Tetrahedron Lett., 1984, 25, 3263.
(c) B. M. Trost and J. D. Oslob, J. Am. Chem. Soc., 1999, 121, 3057.
51 For studies on a,b-epoxy vinyl triflates: K. Yamamoto and C. H. Heathcock,
Org. Lett., 2000, 2, 1709.
52 S. R. Chemler, D. Trauner and S. J. Danishefsky, Angew. Chem. Int. Ed., 2001,
40, 4544.
53 Some related examples have been described: R. Takeuchi and K. Tanabe,
Angew. Chem. Int. Ed., 2000, 39, 1975.
Carbohydr. Chem., 2012, 38, 376–397 | 397

Synthesis of carbohydrate-based artificial
siderophores and their biological
applications
Marta M. Andrade* and Amélia P. Rauter
DOI: 10.1039/9781849734769-00398
During the last decade a series of synthetic siderophores based on mono-

saccharides has emerged. Such compounds bear ligands of the catecholate
and hydroxamate (normal and retro) family and were built around methyl
a-D-glucopyranoside, methyl a-D-galactopyranoside, methyl a-D-ribopyr-
anoside and methyl a-D-xylopyranoside. Studies in Gram-negative bacteria
and mycobacteria have proven that they can act as effective artificial
siderophores.
1 Introduction
Iron is an essential element for microorganisms, plants and animals due to
its exclusive chemical properties, such as the ability to coordinate and
activate oxygen and the possession of ideal redox chemistry for involvement
in electron transport and metabolic processes. Although it is one of the most
abundant elements on the planet, iron(III) solubility at neutral pH is around
10 10M, a value that is four orders of magnitude lower than the con-
centration required by microorganisms and therefore it cannot be utilized
by them.1,2 Consequently, microorganisms, fungi and plants have evolved
strategies to scavenge and absorb iron from soil, fresh and marine water,
and living organisms (plants and animals), by producing high-affinity iron-
binding compounds known as siderophores, capable of chelating iron by
allowing its assimilation through cell surface receptors. Siderophores are
low-molecular-weight compounds (500–1500 dalton) which can dissolve
these ions as soluble Fe3 þ octahedral complexes with high stability con-
stants (log b W 30) that can be taken up by active transport mechanisms
(Table 1).1,3–5 Up to date, there are almost 500 compounds identified as
siderophores.1
Siderophore research stimulates a multidisciplinary effort in order to
facilitate detailed understanding of the essential role of iron in microbial
virulence as well as the potential for the development of effective anti-
microbial agents.6
These compounds can be generally classified according to their iron
binding moieties as catechol-based (enterobactin, agrobactin), hydroxamic
acid-based (ferrichrome, ferrioxamine B), alpha-hydroxycarboxylic acid-
based (vibrioferrin) and mixed-ligand siderophores (aerobactin) (Fig. 1).
Their architecture includes linear, tripodal, endocyclic and exocyclic
Faculdade de Cieˆncias da Universidade de Lisboa, Carbohydrate Chemistry Group, Centre of

Chemistry and Biochemistry/Department of Chemistry and Biochemistry, Edifıćio C8, Piso 5,
Campo Grande, 1749-016 Lisboa, Portugal. E-mail: marta.m.s.andrade@gmail.com
398 | Carbohydr. Chem., 2012, 38, 398–415

Table 1 Relevant thermodynamic parameters for selected siderophore and synthetic side-
rophore complexes.1
Siderophore Log ba pFeb
Enterobactin 49 35.5
Pyoverdin 30.8 27
Ferrichrome A 32.0 25.2
Ferrioxamine E 32.5 27.7
Ferrioxamine B 30.6 26.6
Aerobactin 22.5 23.3
Alcaligin 64.6 23.0
Rhodotorulic acid 62.3 21.8
Acetohydroxamic acid 28.29 12.5
N-Methylacetohydroxamic acid 29.4 16.2
L-Lysinehydroxamic acid 16.1 7.1
Tiron 46.3 20.94
a
Overall stability constant at 25 1C for complexation of Fe(H2O)63 þ by the fully deprotonated
ligand; represents b110 for the hexadentate ligands, b230 for the tetradentate ligands and
b130 for bidentate ligands.
b
log[uncomplexed Fe3 þ aq] at pH 7.4, 1 mM total Fe3 þ and 10 mM total ligand.
structures. The most common natural siderophores are enterobactin, fer-

richrome, coprogen, among others.5,7–9
The siderophore excreted by Streptomyces pilosus, deferrioxiamine B (3,
Fig. 1 and Fig. 2), is commercialized as a mesylated ester named Desferals,
and it was for many years the single drug available in the market for iron
removal in iron overload diseases. However, its retention time in the body is
quite short; its oral activity is very poor, the treatment is expensive and has
several collateral effects, illustrating the importance in developing new
analogs of the drug. Deferiprone was introduced in the late 80s; it is orally
active but does not remove enough iron to keep patients in a negative iron
balance (Fig. 2). In Nov 2005, Deferasirox (Fig. 2) was approved by the
FDA; it is also orally active, has a long-half-life time and some minor side
effects.10 Also a wide range of catecholates have appeared as possible drugs
for iron removal.11,12
Several studies of artificial siderophores have been reported during the last
two decades.3,5,13–20 These compounds attracted growing interest as clinically
useful iron chelators for the treatment of iron overload diseases and for iron
transport-mediated antibiotics or antimicrobial compounds;5 they are also
useful antimalarial agents, acting by causing iron deprivation of malaria
parasites.21 The most common moieties used in artificial siderophores are
catecholate and hydroxamate groups. The stereochemistry of these molecules
is such that the coordination sphere of iron(III) is completely occupied by the
oxygen-containing ligands, resulting in octahedral complexes (Fig. 3).18
The development of artificial siderophores that can act as functional
model compounds for native siderophores is an essential tool for the
investigation of siderophore mechanisms of recognition and metabolism.
Molecular recognition and signalling are at the heart of biological pro-
cesses, by governing intercellular communication and information transfer.
The recognition phenomena of siderophore mediated microbial iron(III)
uptake involves two processes: recognition of iron(III) by microbial
Carbohydr. Chem., 2012, 38, 398–415 | 399

Fig. 1 Examples of siderophores classified according to their iron binding moieties (all shown
without iron). The atoms in bold represent the binding sites.
siderophores as iron(III) carriers and recognition and binding of side-

rophore-iron(III) complexes by microbial membrane receptors. Conse-
quently, many efforts are nowadays devoted to the design and synthesis of
less toxic analogs and, in this context, the use of siderophores with carbo-
hydrate skeleton is a promising subject to be exploited.
2 Artificial siderophores - background

Exocyclic and tripodal siderophore mimics have been successfully built
around tertiary amine,17,19,22 benzene,23,24 tetrahedral carbon,25–28 lysine
derivatives20,29 or cyclitols,30–32 maintaining the chelating capacities but
sacrificing its water solubility and the receptor-recognition capability of the
400 | Carbohydr. Chem., 2012, 38, 398–415

Fig. 2 Drugs used nowadays for the treatment of iron overload diseases (binding atoms in
bold).
Fig. 3 Schematic representation of chelate ring formation for Fe(III)/siderophore complexes.18
backbone. Thermodynamically, these synthetic analogues have Fe(III)

chelating capabilities that are comparable if not greater than those of their
natural counterparts. Inomata et al. have also demonstrated that the bulky
terminal groups and intramolecular hydrogen bonds on artificial side-
rophores play important roles in recognition and permeation.28
Raymond et al.33 reported the total synthesis of naturally occurring
siderophores and Neilands and Miller34,35 have synthesized various artificial
siderophores with binding constants greater than 1020 M1; however, larger
binding constants have been obtained by Vögtle and coworkers.36 More
recently, analogues of siderophores with cyclic monosaccharide backbone
Carbohydr. Chem., 2012, 38, 398–415 | 401

were synthesized and characterized, showing Fe(III) chelating capabilities
that are comparable, if not greater than those of their natural counterparts;
biologically, they are highly promising synthetic analogues, as in vivo studies
show significant siderophore activity indicating favorable cell receptor
recognition and cellular uptake.13–16,37,38 The design of siderophore mimics
requires some important features to be considered, which include the nature
of donor groups, denticity, chelate ring size, architecture, hydrophobicity/
hydrophilicity and stability. The availability of different stereoisomeric forms
of saccharides and their high functionality render them particularly attrac-
tive for the conception of siderophore model compounds of varying structure
and polarity and the construction of an optimal octahedral binding cavity
for the Fe(III) ion. The chiral saccharide backbone can provide favorable
recognition and accessibility to cells, high hydrolytic stability and increased
hydrophilicity. The presence of oxygen atoms in and around the saccharide
ring gives the cyclic backbone the ability to act as a hydrogen-bond-acceptor
during the cell receptor-recognition process. Such compounds and their
physicochemical properties are of interest as they can be used as penetration
vectors for antibiotics or other biological applications involving Fe(III)
uptake and metabolism. These results suggest that the carbohydrate scaffold
is a promising biologically viable backbone for siderophore analogues.
3 Synthesis of Carbohydrate-based artificial siderophores

3.1 Catecholate derivatives
Catechol moieties show high affinity for iron(III) as they interact strongly
with tripositive metal cations, resulting in high electron density of both
oxygen atoms. The binding of cations by catechol has marked pH sensi-
tivity, and the complexes forming at pH 7.0 each bear a net charge and
consequently are unlikely to permeate membranes by simple diffusion, thus
tending to be trapped in intracellular compartments. A disadvantage for
catechol-based ligands is their susceptibility towards oxidation.18
The most relevant natural catecholate siderophore is enterobactin (1,
Fig. 1), which is produced by E. coli and other enteric bacteria under iron
limitation. Studies indicate that the chiral trilactone backbone and the spacer
chain length are to some extent responsible for the high stability of the
Fe(III)-enterobactin complex and are critical in the receptor mediated
transport and iron release mechanism. In 2001, Heggemann et al. have shown
that carbohydrates can substitute advantageously the trilactone structure of
enterobactin.14,16 In the last decade, the synthesis of carbohydrate-based
artificial siderophores was focused only on monosaccharide scaffolds. The
group synthesized several glucose derivatives with two or three 2,3-di(ace-
toxy)benzoyl ligands at secondary carbon positions of the glucopyranose ring
and also with three 2,3-di(acetoxy)phenoxyacetyl and 8-acetoxy-2-oxo-3,4-
dihydro-2H-benzo[1,3]oxazin-3-yl)acetyl ligands at the same positions. The
synthesis of biscatecholate derivatives started from methyl 4,6-O-iso-
propylidene-a-D-glucopyranoside (7), which has two free hydroxyl groups
ready to react with 2,3-di(acetoxy)benzoyl chloride (8) in basic conditions,
using THF as solvent. Posterior removal of isopropylidene protecting group
afforded biscatecholate 10 with overall yield of 58% (Scheme 1).
402 | Carbohydr. Chem., 2012, 38, 398–415

Scheme 1 Synthesis of bis-catecholate derivatives with di(acetoxy)benzoyl ligands at carbon
positions 2 and 3.16
Scheme 2 Synthesis of triscatecholate derivatives with different ligands at carbon positions 2,

3 and 4.16
The approach to obtain trischatecolate siderophores started from 6-O-

trityl protected methyl a-D-glucopyranoside derivative (11), which has three
free hydroxyl groups at carbon positions 2, 3 and 4. A spacer was now
introduced between the sugar and the ligand via a Michael type cya-
noethylation with acrylonitrile, followed by reduction of the nitrile with
CoCl2/NaBH4 to afford ligand 13 (Scheme 2). Different chatechol-based
chelating groups were then introduced in the molecule, e.g. by reaction with
the acyl chlorides 14a–d. Final removal of the trityl group with boron tri-
fluoride etherate afforded the triscatecholate siderophores 15a–d. Com-
pounds 14a and 14d were also deacetylated with NaOH to afford 16 and 17
with free catechol-OH groups (Scheme 3).
Also Dhungana et al.14 have reported free-catecholate derivatives, e.g.
starting from 15a and 17 (Scheme 4).
Another series of triscatecholates based on methyl a-D-glucopyranoside
and methyl a-D-galactopyranoside (Schemes 5 and 6) was also reported
using synthetic pathways similar to those previously described.15
New dihydroxybenzylidenehydrazino derivatives based on methyl a-D-
glucopyranoside by replacing the aminopropyl groups by hydrazinocarbo-
nylethyl functional groups have been reported as a new variation of the
carbohydrate siderophores (Scheme 7).
Carbohydr. Chem., 2012, 38, 398–415 | 403

Scheme 3 Synthesis of a triscatecholate derivative with free catechol-OH groups.16
Scheme 4 Triscatecholate derivatives with free catechol-OH groups.14
3.2 Retro-hydroxamate and hydroxamate derivatives

Hydroxamate ligands possess lower affinity for iron than catecholates but
they have the advantage of forming neutral tris-complexes with iron(III),
which are in principle capable of permeating membranes by nonfacilitated
diffusion. Many hydroxamates are metabolically labile and are only poorly
absorbed via the oral route.18
Desferrioxamine B (3, Fig. 1) is a hydroxamate-type siderephore used
in medicine to treat potentially fatal iron overload resulting from regular
blood transfusion of patients with thalassaemia, a genetic blood disease
404 | Carbohydr. Chem., 2012, 38, 398–415

Scheme 5 Synthesis of new triscatecholate derivatives with methyl a-D-glucopyranoside
backbone.15
Scheme 6 Synthesis of a triscatecholate derivative with methyl a-D-galactopyranoside

backbone.15
Scheme 7 Synthesis of dihydroxybenzylidenehydrazinocatecholates based on methyl a-D-

glucopyranoside.15
characterised by the decreased production of normal haemoglobin and

resulting in anaemia. Desferichrome (4, Fig. 1) is another example of a
natural tris-hydroxamate siderophore with an exocyclic structure; its iron
(III) complex, called ferrichrome was one of the first siderophores
Carbohydr. Chem., 2012, 38, 398–415 | 405

discovered in nature from the smut fungus, Ustilago sphaerogena, and is also
produced by fungi of the genera Aspergillus, Ustilago and Penicillium
among others.39 Retro-hydroxamate siderophores are analogues in which
the position of the hydroxamate nitrogen and carbon are interchanged
relative to their position in the natural siderophore and have found appli-
cation as antimalarial agents.40–43
The first synthetic analogues of hydroxamate siderophores based on
methyl a-D-glucopyranoside, methyl a-D-galactopyranoside, methyl a-D-
ribopyranoside and methyl a-D-xylopyranoside were in fact retro-hydro-
xamates and were reported by Heggemann et al.15 The two hexoses were
first protected with a trityl group at C-6, so that all four monosaccharides
studied have only three free OH groups (OH-2, 3 and 4). The corresponding
retro-trishydroxamates with different spacer groups were prepared by
Steglich esterification with N-benzoyloxy-N-methylglutaric acid (30–33a),
N-benzyloxy-N-methylsuccinic acid (30–33b), N-benzyloxy-N-methylglu-
taric acid (30–33c) and N-benzyloxy-N-butylglutaric acid (30d), using
dicyclohexyl-carbodiimide (DCC) and N,N-dimethylaminopyridine
(DMAP) in dichloromethane. Removal of trityl protecting groups was
accomplished for gluco and galacto derivatives with boron trifluoride
etherate, and debenzylation by catalytic hydrogenation afforded com-
pounds 32–35e,f and 34g with free hydroxamate groups (Scheme 8).
In 2008, Gebhardt et al.38 synthesized new glucose-based retro-hydro-
xamates, with different N-alkyl groups by using the same methods described
Scheme 8 Synthesis of retro-hydroxamate derivatives with hexoses and pentoses.15
406 | Carbohydr. Chem., 2012, 38, 398–415

Scheme 9 Synthesis of glucose-based retro-hydroxamates with long N-alkyl chains.38
Scheme 10 Synthesis of normal-type hydroxamates based on glucose and mannose.13,38
before. Surprisingly, the authors observed the unequivocal formation of a

2,3,6-substituted glucopyranoside (37) as by-product, probably due to a
Lewis acid catalyzed migration (Scheme 9).
More recently, Dhungana13 and Gebhardt38 reported the first synthetic
analogues of desferrichrome with normal-hydroxamate geometry based on
methyl a-D-glucopyranoside and methyl a-D-mannopyranoside with three
asymmetric hydroxamic acid moieties at carbon positions 2, 3 and 4 (38,
40), or 2, 3 and 6 (39) (Scheme 10).
4 Study of siderophore activity

In Gram-negative bacteria the Fe(III)-siderophore complex is first captured
at the cell surface by specific protein receptors, followed by transport to the
periplasmic space. For E. coli, the most common bacterial strain used in the
laboratory, siderophore-iron complexes are recognized by outer membrane
receptors which are produced by microbes under iron stressed conditions.
Different strains of bacteria produce different types of siderophore recep-
tors. For example, E. coli expresses various outer membrane proteins
including the ferrichrome receptor (FhuA) and enterobactin receptor
(FepA) while Pseudomonas putida produces pseudobactin 358 receptor
(PupA).5 The siderophore activity of all siderophore analogs was examined
Carbohydr. Chem., 2012, 38, 398–415 | 407

by growth promotion assays with various Gram-negative bacteria and
mycobacteria, which have been well defined in their ability to transport and
utilize natural siderophores (siderophore indicator strains).
Gram-negative indicator strains (Iron related marker) included Salmo-
nella typhimurium enb-7 (enterobactin-) and TA 2700 (enterobactin-, FEB
C-), E. coli AB 2847 (aromates biosynthesis-), E. coli IR 112 (aroB-, tonB-),
Klebsiella pneumoniae KN 4401(enterobactin-, aerobactin biosynthesis-),
Morganella morganii SBK 3 (wild type),Yersinia enterocolitica H 5030
(yersiniabactin-) and Pseudomonas aeruginosa PAO 6609 (pyoverdin-).16
Compounds 9, 10, 14a–d, 15a–d, 16 and 17 did not stimulated the growth
of the indicator strains P. aeruginosa PAO 6609, Klebsiella pneumoniae KN
4401 and E. coli IR 112 missing the protein tonB. With this result the
authors suggested that the presence of tonB is necessary for siderophore
activity of these compounds. With other bacterial strains, different results
were obtained. Biscatechol derivatives 9 and 10 displayed only a low
growth promotion of the indicator strains S. typhimurium enb-7 and
E. coli AB 2847. Triscatecholate 15a promoted very strongly the growth of
indicator strains S. typhimurium enb-7, TA 2700 and E. coli AB 2847 but
not that of the other indicator strains; derivative 15b showed only a minor
activity against the strains P. aeruginosa ATCC 27853, SG 137 and E. coli
ATCC 25922. However, compound 15c promoted Yersinia enterocolitica
H 5030. Trityl substituted derivatives (14a–d, 16 and 17) were of lower
activity.16
Studies with Gram-negative strains of the antibacterial screening, grown
under iron limitation, with P. aeruginosa ATCC 27853, SG 137, NCTC
10662 and E. coli ATCC 25922 were also accomplished. These experiments
showed that compounds 15a–d and 17 were active in growth promotion of
the strains, while compounds 14a,b and 16 were less active and no activity
was found for biscatecholates 9 and 10.16
The previous compounds were also studied for growth promotion of
mycobacteria, with wild type strain of M. smegmatis SG 987 and mc2155,
mutants of M. smegmatis M10 (exochelin-) and M24 (mycobactin-), a wild
type of M. phlei 239 (exochelin-) and the mutant B1 of M. smegmatis
mc2155 generated by gene replacement (blocked in exochelin
biosynthesis).16
Here again, bischatecolates were inactive and the most active substances
were 17 (the derivative with elongated spacer) and 15a. The remaining
compounds showed diverse results.16
Finally, the compounds were tested for their iron complexing capacity
using the chromazurol-S test (CAS-assay) and 15a, 15b, 15d and 17 were
active.16
The complete set of results for siderophore activity of compounds 9, 10,
14a–d, 15a–d, 16 and 17 can be found in Tables 2 and 3.
Free trishydroxamates 32e, 33d, 33f, 34d, 34e, 34g and 35f were active on
all test strains using wild type strains of Gram-negative bacteria under iron
limitation and the enterobactin mutant S. typhimurium enb 7. Galactose
derivative 35f, with longer spacer group was more active than 35e; however,
the glucose derivative with shorter spacer (34d) was more active than the
one with longer spacer (34e). These results can indicate that the siderophore
408 | Carbohydr. Chem., 2012, 38, 398–415

Table 2 Growth promotion tests of the synthesized siderophore analogues on Gram-negative
bacteria. Diameter of growth zone (mm), substance application 5 mg.16
Compound no. 9 10 14a 14b 14c 14d 15a 15b 15c 15d 16 17 control
Indicator strain
S. typhimurium:enb-7 10 17 15 0 0 18 32 0 0 0 0 18 28a
S. typhimurium TA 2700 0 0 10 n.d. n.d. n.d. 33 n.d. n.d. n.d. n.d. n.d. 30b
E. coli AB 2847 0 10 0 0 0 0 25 0 0 0 0 0 28c
M. morganii SBK 3 0 0 0 0 0 0 10 0 0 24 0 10 22d
Y. enterocolitica H5030 0 0 0 0 10 0 0 0 16 0 10 0 18c
Strains of antibacterial screening
P. aeruginosa ATCC 27853 0 0 0 0 0 0 n.d. 0 15 14 0 12 35e
P. aeruginosa SG 137 0 0 15 16 15 11 22 16 16 14 0 15 30e
P. aeruginosa NCTC 10662 0 0 15 0 0 14 25 15 22 17 0 15 30e
P. aeruginosa ATCC 9027 0 0 0 0 15 0 25 0 17 18 0 15 35e
E. coli ATCC25922 0 0 12 0 15 15 35 12 22 21 11 19 38f
a
ferrioxamin G.
b
enterobactin.
c
ferrichrome.
d
2,3-dihydroxybenzylidene-1,3,5-trimethylalinine.
e
desferal.
f
ferricrocin (2 mg).
Table 3 Growth promotion tests of the synthesized siderophore analogues on strains of

mycobacteria. Diameter of growth zone (mm), substance application 5 mg.16
Mycobactin
Compound no. 9 10 14a 14b 14c 14d 15a 15b 15c 15d 16 17 (2 mg)
Indicator strain
SG 987 0 0 9 0 13 10 23 0 10 10 14 28 17
M10 0 0 16 13 14 13 21 0 0 0 15 20 17
M77 0 0 0 0 0 0 0 10 0 0 0 0 17
mc2 155 0 0 15 0 16 12 25 13 14 14 15 30 19
M24 0 0 0 0 0 0 26 0 0 0 0 8 19
B1 0 0 10 10 13 11 20 17 15 15 13 25 16
activity results from a joint influence of the length of the spacer groups and
of the individual sugar backbone component.15 The complete set of results
can be observed in Table 4.
The results obtained on mycobacteria showed that compounds with free
OH groups have siderophore activity as well as effective iron complexation,
while 20 did not complexe with iron. Compound 21 was also active on strain
mc2155 and B1. The derivatives 27a (with 2,3-OH groups) and 27b (with
methoxycarbonyloxy groups) exhibited preferred activity on P. aeruginosa
strains. Compound 34g bearing a lipophilic butyl component was active on
the wild type strain mc2155 and on the mutant B1. Iron complexes of the
free hydroxamates 30e, 33d, 33f, 34d, 34e, 34g, 35e and 35f, as well as
the iron complexes of all triscatecholates were effective growth promoters of
the wild type strain mc2155 and the single mycobactin (M24) or exochelin
(B1) mutant (Table 5).15
Carbohydr. Chem., 2012, 38, 398–415 | 409

Table 4 Growth promotion tests of Gram-negative bacteria, diameter of growth zone (mm), substance application 10 mmol on 6 mm paper discs.15
Compound no. 20 21 25 27a 27b 30a 31a 32a 32e 33a 33d 33f 34a 34d 34e 34g 35a 35e 35f control
Indicator strain
P. aeruginosa 14 25 20 25 16 0 0 0 24 0 17 20 0 18 11 22 0 0 21 35a
ATCC 27853
P. aeruginosa SG 137 12 21 23 23 18 0 0 0 24 0 25 24 0 20 20 20 0 19 25 30a
P. aeruginosa NCTC 10662 11 28 25 25 17 0 0 0 24 0 24 29 0 25 19 20 11 14 28 31a
P. aeruginosa ATCC 9027 10 n.d. n.d. 22 0 0 0 0 22 0 20 26 0 19 14 20 0 15 26 34a
410 | Carbohydr. Chem., 2012, 38, 398–415

P. aeruginosa K799/WT 15 23 23 20 15 0 0 0 18 0 25 28 0 20 11 22 0 10 25 35a
E. coli ATCC25922 0 27 26 15 0 0 0 0 26 0 23 27 13 22 16 23 0 17 19 33b
S. typhimurium enb-7 0 20 26 14 0 0 0 0 17 0 17 27 0 12 10 19 0 0 28 32a
a
desferal G.
b
ferricrocin.
Table 5 Growth promotion of mycobacteria, diameter of growth zone (mm), substance application 10 mmol (mycobactin 2 mg) on 6 mm paper discs (with and without iron
complexation).15
Compound no. 20 21 25 27a 27b 30a 31a 32a 32e 33a 33d 33f 34a 34d 34e 34g 35a 35e 35f A B
Indicator strain
mc2 155 10 24 22 0 0 0 0 0 0 8 0 0 0 0 0 25 9 0 0 - -
mc2 155 þ Fe - 30 27 29 26 - - - 33 - 30 34 - 18 24 28 - 22 36 30 14
M24 0 0 0 0 0 0 0 8 0 0 0 0 0 0 0 0 0 0 0 - -
M24 þ Fe - 18 16 20 21 - - - 32 - 24 31 - 22 27 19 - 15 33 28 16
B1 0 21 22 0 0 0 0 0 0 0 0 18 7 0 0 26 0 0 0 - -
B1 þ Fe - 32 23 26 21 - - - 30 - 25 33 - 21 27 25 - 21 36 32 15
B3 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 - -
B3 þ Fe - 14 0 0 0 - - - 19 - 18 15 - 15 12 13 - 13 28 25 15
U3 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 - -
U3 þ Fe - 13 0 0 0 - - - 24 - 20 17 - 16 12 12 - 0 26 22 16
A: ferrichrome; B: mycobactin.
Carbohydr. Chem., 2012, 38, 398–415 | 411

Table 6 Growth promotion of E. coli wild type strains and mutants. Diameter of growth zone
(mm), substance application 10 mmol.15
Compound no. 16 17 21 25 33f 35f 2,3-DHBAa Ferricrocin
Indicator strain
AB 2847 32 25 25 29 24 23 30 30
BR 158 0 0 0 0 0 0 0 0
H 1443 32 26 25 31 25 23 32 30
H 18765 0 0 0 0 26 24 0 30
H 873 27 25 24 27 25 21 30 31
H 1877 25 25 21 25 27 22 27 32
H a875 25 22 22 25 27 23 30 33
a
2,3—DHBA=2,3-dihydroxybenzoic acid.
The recognition and uptake specificity of the siderophores by use of

E. coli K12 mutants with alterations in the TonB protein was studied for
relevant compounds using ferricrocin as control (Table 6).
Siderophore activity of compounds 36-40 (Table 7) was tested using
M. smegmatis wild type strains and mutants in siderophore biosynthesis and
transport.38 The authors observed stronger growth promotion activity for
iron complexed siderophores compared to their iron free analogs. The
hydroxamates 38–40 have shown nearly zero activity, but when complexed
with iron strong growth promotion activity was detected. In contrast, retro-
hydroxamates 36a, 36b and 37 promoted growth of the mycobacteria also in
both ferri/desferri forms. These results suggest that the growth promotion
activity might depend on lipophilicity rather than on siderophore receptor
mediated uptake, since the higher lipophilic character of the retro-hydro-
xamates facilitates diffusion into the hydrophobic mycobacterial cell
envelope.38 It was observed that glucose-based ferri-hydroxamates 38 and
39 have improved growth promoting activity for both of the mutant strains
mc2155-M24-B3 and mc2155-M24-U3, which are completely blocked in
their species specific exochelin/mycobactin iron uptake system. Mannose-
based ferri-hydroxamate 40 and glucose-based ferri-retrohydroxamates 36a,
36b and 37 were less effective for exochelin/mycobactin biosynthesis mutant
mc2155-M24-B3 and inactive for the mycobactin biosynthesis/exochelin
uptake mutant mc2155-M24-U3 (Table 7).38
5 Conclusions and perspectives

During the last decade a series of synthetic analogs of bacterial siderophores
built around a carbohydrate skeleton were reported in the literature. These
new carbohydrate-based artificial siderophores were focused only on
monosaccharide scaffolds. The results obtained indicate that such com-
pounds are highly promising synthetic analogs, as in vivo studies showed
significant siderophore activity, indicating favorable cell receptor recogni-
tion and cellular uptake. In this context, the use of siderophores with car-
bohydrate skeleton merits to be further exploited. The use of disaccharide
platforms (such as sucrose) with several possible hydroxyl groups to explore
as attachment point for the ligands would allow to study the importance of
412 | Carbohydr. Chem., 2012, 38, 398–415

Table 7 Growth promotion of mycobacteria with Mycobacterium smegmatis strains.38
Compound no. Aa Ba Cb 36a 36a þ Fe 36b 36b þ Fe 37 37 þ Fe 38 38 þ Fe 39 39 þ Fe 40 40 þ Fe D E
Indicator strain
SG 987 þ þ þ þþþþ þþþþ þþþþ þþþ þþþþ þþþþ þ þþþþ þ þþþ þ þþþ þþ þþ
SG 987–M10 þ þ þþ þþþ þþþ þþþ þþ þþþ þþþ þþþþ þþþ þþ þþ
mc2 155 þ þ þ þþþþ þþþþ þþþþ þþþ þþþþ þþþþ þ þþþþ þþþþ þ þþþþ þþ þþ
mc2 155–M24 þ þ þþþþ þþ þþþ þþþþ þþþþ þþþ þþ þþ
mc2 155–B1 þ þ þþþ þþþþ þþþþ þþ þþþþ þþþþ þ þþþþ þþþþ þ þþþþ þþ þ
mc2 155–M24–B3 þ þþ þ þþ þþ þþþ þ þþ
mc2 155–M24–U3 þ þ þ þ þþ þþþ þþ
a
indicates that the strain biosinthesizes the indicated siderophore and has the potential for ligand exchange, yes(þ) or no().
b
uptake pf exochelin by the indicated strains, yes( þ ) or no( ).
c
diameter of orange halo in mm , þ , þþ , þþþ indicates the presence and increasing intensity of growth promoting activity.
Carbohydr. Chem., 2012, 38, 398–415 | 413

the position of the ligands in the saccharide backbone and also to evaluate
siderophores with different hydrophilicity/lipophilicity. Other promising
applications in this area may arise by attaching a fluorescent probe with a
linker to the artificial siderophore, which will not interfere with iron che-
lation and subsequent receptor recognition, and that would make possible
to develop new iron-sensitive fluorescent sensors for biological signaling
and biomedical or environmental applications.
References
1 H. Boukhalfa and A. L. Crumbliss, BioMetals, 2002, 15, 325–339.
2 V. Braun and H. Killmann, Trends Biochem. Sci., 1999, 24, 104–109.
3 K. Matsumoto, T. Ozawa, K. Jitsukawa and H. Masuda, Inorg. Chem., 2004,
43, 8538–8546.
4 B. F. Matzanke, in Encyclopedia of Inorganic Chemistry, John Wiley & Sons,
Ltd, 2006.
5 J. M. Roosenberg, Y. M. Lin, Y. Lu and M. J. Miller, Curr. Med. Chem., 2000,
7, 159–197.
6 G. M. Rodriguez, R. Gardner, N. Kaur and O. Phanstiel, BioMetals, 2008, 21,
93–103.
7 Iron transport in bacteria, ASM Press, Washington D.C., 2004.
8 A. M. Albrecht-Gary and A. L. Crumbliss, Metal ions in biological systems,
New York, 1988.
9 A. L. Crumbliss, Handbook of microbial iron chelates, CRL Press, Boca Ranton,
FL, 1992.
10 E. J. Neufeld, Blood, 2006, 107, 3436–3441.
11 R. J. Bergeron, C. Z. Liu, J. S. McManis, M. X. B. Xia, S. E. Algee and J.
Wiegand, J. Med. Chem., 1994, 37, 1411–1417.
12 I. Turcot, A. Stintzi, J. D. Xu and K. N. Raymond, J. Biol. Inorg. Chem., 2000,
5, 634–641.
13 S. Dhungana, J. M. Harrington, P. Geblhardt, U. Mollmann and A. L.
Crumbliss, Inorg. Chem., 2007, 46, 8362–8371.
14 S. Dhungana, S. Heggemann, L. Heinisch, U. Mollmann, H. Boukhalfa and
A. L. Crumbliss, Inorg. Chem., 2001, 40, 7079–7086.
15 S. Heggemann, U. Mollmann, P. Gebhardt and L. Heinisch, BioMetals, 2003,
16, 539–551.
16 S. Heggemann, M. Schnabelrauch, D. Klemm, U. Mollmann, R. Reissbrodt
and L. Heinisch, BioMetals, 2001, 14, 1–11.
17 L. Heinisch, P. Gebhardt, R. Heidersbach, R. Reissbrodt and U. Möllmann,
BioMetals, 2002, 15, 133–144.
18 Z. D. Liu and R. C. Hider, Coord. Chem. Rev., 2002, 232, 151–171.
19 K. Matsumoto, N. Suzuki, T. Ozawa, K. Jitsukawa and H. Masuda, Eur. J.
Inorg. Chem., 2001, 2481–2484.
20 S. Wittmann, L. Heinisch, I. Scherlitz-Hofmann, T. Stoiber, D. Ankel-Fuchs
and U. Möllmann, BioMetals, 2004, 17, 53–64.
21 A. Tsafack, J. Libman, A. Shanzer and Z. I. Cabantchik, Antimicrob. Agents
Ch., 1996, 40, 2160–2166.
22 C. Y. Ng, S. J. Rodgers and K. N. Raymond, Inorg. Chem., 1989, 28, 2062–
2066.
23 T. Masuda, S. Asari, S. Tamagaki and T. Nagasaki, Tetrahedron Lett., 2000, 41,
2411–2414.
24 I. Yoshida, I. Murase, R. J. Motekaitis and A. E. Martell, Can. J. Chem., 1983,
61, 2740–2744.
414 | Carbohydr. Chem., 2012, 38, 398–415

25 M. M. Meijler, R. Arad-Yellin, Z. I. Cabantchik and A. Shanzer, J. Am. Chem.
Soc., 2002, 124, 12666–12667.
26 H. Weizman, O. Ardon, B. Mester, J. Libman, O. Dwir, Y. Hadar, Y. Chen and
A. Shanzer, J. Am. Chem. Soc., 1996, 118, 12368–12375.
27 H. Ouchetto, M. Dias, R. Mornet, E. Lesuisse and J. M. Camadro, Bioorg.
Med. Chem., 2005, 13, 1799–1803.
28 T. Inomata, H. Eguchi, K. Matsumoto, Y. Funahashi, T. Ozawa and H.
Masuda, Biosens. Bioelectron., 2007, 23, 751–755.
29 Y. M. Lin and M. J. Miller, J. Org. Chem., 1999, 64, 7451–7458.
30 A. W. Coleman, C. C. Ling and M. Miocque, Angew. Chem.-Int. Edit. Engl.,
1992, 31, 1381–1383.
31 T. Masuda, J. Hayashi and S. Tamagaki, J. Chem. Soc.-Perkin Trans., 2000, 2,
161–167.
32 B. Tse and Y. Kishi, J. Am. Chem. Soc., 1993, 115, 7892–7893.
33 K. N. Raymond and C. J. Carrano, Acc. Chem. Res., 1979, 12, 183–190.
34 B. H. Lee, M. J. Miller, C. A. Prody and J. B. Neilands, J. Med. Chem., 1985,
28, 317–323.
35 B. H. Lee, M. J. Miller, C. A. Prody and J. B. Neilands, J. Med. Chem., 1985,
28, 323–327.
36 W. Kiggen and F. Vögtle, Angew. Chem., Int. Ed., 1984, 23, 714–715.
37 S. Dhungana, S. Heggemann, P. Gebhardt, U. Mollmann and A. L. Crumbliss,
Inorg. Chem., 2003, 42, 42–50.
38 P. Gebhardt, A. L. Crumbliss, M. J. Miller and U. Mollmann, BioMetals, 2008,
21, 41–51.
39 A. Butler and R. M. Theisen, Coordin. Chem. Rev., 2010, 254, 288–296.
40 J. Golenser, A. Tsafack, Y. Amichai, J. Libman, A. Shanzer and Z. Cabantchik,
Antimicrob. Agents Chemother., 1995, 39, 61–65.
41 K. P. Holland, H. L. Elford, V. Bracchi, C. G. Annis, S. M. Schuster and D.
Chakrabarti, Antimicrob. Agents Chemother., 1998, 42, 2456–2458.
42 M. J. Miller, A. J. Walz, H. Zhu, C. Wu, G. Moraski, U. Möllmann, E. M.
Tristani, A. L. Crumbliss, M. T. Ferdig, L. Checkley, R. L. Edwards and H. I.
Boshoff, J. Am. Chem. Soc., 2011, 133, 2076–2079.
43 P. K. Ojha and K. Roy, Mol. Simul., 2010, 36, 939–952.
Carbohydr. Chem., 2012, 38, 398–415 | 415

Smart biomaterials: the contribution of
glycoscience
Laura Cipolla,* Laura Russo, Francesca Taraballi, Cristina Lupo,
Davide Bini, Luca Gabrielli, Alice Capitoli and Francesco Nicotra
DOI: 10.1039/9781849734769-00416
Examples of material functionalisation (‘‘biodecoration’’) with signalling

and relevant glycidic scaffolds will be outlined. Recent research concerning
the development of smart biomaterials for Tissue Engineering (TE) appli-
cations will be considered.
1 Introduction: biomaterials and tissue engineering

Humankind’s use of materials to repair the body dates to antiquity, when
natural materials such as wood were used in an attempt to structurally
replace injured tissues by diseases or trauma. In the beginning of the
twentieth century, synthetic polymers, ceramics and metal alloys were
introduced in place of natural materials; they offer better performance,
increased functionality and more reproducibility than their naturally
derived counterparts. These advances led to a pronounced increase in the
range of use and the efficacy of biomaterials, as a result of which millions of
lives have been saved or improved by devices such as vascular stents, dental
restoratives, artificial hips and contact lenses. The increasing importance of
biomaterials in our society over the past decades can be tracked in a number
of ways, including the growth of tissue engineering applications both as an
academic discipline and as an important industry.1
Sales of regenerative biomaterials already exceed US$240 million per
annum2 and the wider markets that tissue engineering taps into are colossal:
costs related to organ replacement account for 8% of global healthcare
spending, and by 2040 as much as 25% of the US GDP is expected to be
related to healthcare.3 Nevertheless, if the short history of industrial tissue
engineering has taught us anything, it is that the provision of effective
products is not in itself sufficient to ensure commercial success. Early TE
efforts were plagued by product issues related to scale-up, shelf-life, quality
control and distribution, and suffered from inappropriate business models
and withdrawal of private finance in the early 2000s.4 Since then the field
has matured, evidenced by the return of large-scale investment and the first
regenerative medicine companies becoming profitable. Alongside these
positive developments, progress in biomaterials design and engineering are
converging to enable a new generation of instructive materials to emerge as
candidates for regenerative medicine.5
Initially, the main goal of scientists was the design of materials largely
performing mechanical functions, being preferably ‘inert’ without interac-
tion with the biology of the host organism in order to prevent rejection.6,7
Dept. of Biotechnology and Biosciences University of Milano-Bicocca, P.za della Scienza 2,

20126 Milano, Italy. E-mail: laura.cipolla@unimib.it
416 | Carbohydr. Chem., 2012, 38, 416–445

Early research and fortuitous accidents linking materials chemistry to bio-
logical response provided a rational basis for developing new biomaterials.1
The molecular biology revolution of the 1970s and advances in genomics,
proteomics and more recently in glycomics significantly affected the ways in
which biomaterials are nowadays designed and used.
It is well known that a series of interactions occur between the surface of
biomaterials and the biological environment after they have been implanted
into the human body. Therefore, the biomaterials surface plays an extre-
mely relevant role in the response of artificial medical devices to the bio-
logical environment.
The efficacy of artificial implants is determined mainly by their surface
characteristics such as surface morphology, microstructure, composition,
and functionalization. These properties alter adsorption of proteins which
mediate the adhesion of desirable and undesirable cells.8 Research on bio-
materials surface has become one of the hottest topics in biomaterials and
biomedical engineering. The highest score for a keyword in the World
Biomaterial Congress in 2008 (WBC 2008) is Biomaterials Surface, indi-
cating that the current research focus of the biomaterials community is
aimed at understanding the fundamental processes at the interface between
implant surfaces and surrounding living tissues.9
To this end, scientists are creating new materials including those with
improved biocompatibility, stealth properties, cell responsiveness (smart
materials), specificity and other critical properties. Modern biomaterials
science is characterized by a growing need to integrate biomaterials design
with new insights emerging from studies of cell–matrix interactions, cellular
signalling processes, developmental and systems biology.10
Since cell contact with the biomaterial surface significantly influences cell
behaviour and performance, trends in biomaterial designs lean towards
bioactive materials that can modulate and control cell behaviour.
1.1 Tissue engineering and the biomimetic approach

A variety of new materials are being synthesized from man-made building
blocks, and being used to create devices for specific medical applications.5,11
Materials composed of naturally occurring (biologically derived) building
blocks, including extracellular matrix (ECM) components, are being studied
for applications such as direct tissue replacement and tissue engineering.
The ECM, a complex composite of proteins, glycoproteins and proteogly-
cans, provides an important model for biomaterials design (Fig. 1).12
ECM-derived macromolecules (for example, collagen) have been used for
many years in biomaterials applications5 and it is now possible to create
artificial analogues of ECM proteins using recombinant DNA technology.13
Tissue engineering is an advanced interdisciplinary field that encompasses
the design of biomaterials for in vivo tissue regeneration where mimicking
the natural ECM is often pursued.
The natural ECM supports organ and tissue structure and function, and
also regulates basic cellular events like proliferation, growth, migration,
differentiation, and survival.14 These functions are controlled through tissue-
specific constituents, such as collagens, laminins, fibronectin or elastins, as
well as functional molecules like growth factors or matricellular proteins,
Carbohydr. Chem., 2012, 38, 416–445 | 417

Fig. 1 ECM-cell interaction.
among others. Novel biomaterials should allow for the gradual endogenous
remodeling of native tissue leading to the replacement of implant material,
manufactured to replace a missing biological structure, with fully functional
ECM and cells that existed at the implant site prior to damage.
The outer membrane of a typical cell is covered by specific carbohydrate
structures and a forest of at least six different receptor systems that can be
activated by interactions with adjacent cells, ligands in the surrounding
ECM, and secreted signalling molecules. Hundreds of different proteins
play a role in the composite stimulation of cell receptors, which in turn
determine a plethora of responses, including cell migration in the early
embryo, coordinated organogenesis, and wound repair throughout adult
life.15–19 Collectively, these extrinsic factors make up a highly defined and
specialized cell microenvironment, which is essential for correct tissue
development and continued function.
The ECM takes a variety of forms and composition in different tissues
and at different stages of development in the same tissue.20 Diversity arises
through combinations of specific molecular interactions between numerous
isoforms, ratios, and geometrical arrangements of collagens, elastins, pro-
teoglycans, and adhesion proteins such as fibronectins and laminins. This
creates an environment that is replete with informational cues. In addition
to this, a wealth of molecular mechanisms modulates the dissemination of
the biological information.21
The critical point during in vitro tissue engineering is to at least partially
recreate conditions that mimic the natural ECM environment for particular
cell types in order to support their function and proliferation. Since cell
contact with the biomaterial surface is a key point, in recent years, bio-
materials design has focused on the exposition and incorporation of sig-
nalling molecules into scaffold materials rather than using them in a
diffusive or soluble way (Fig. 2).22
Among the most studied molecules are multifunctional proteins like
growth factors23–26 or cytokines,27 while there are also reports on use of
small molecules like neurotransmitters (Fig. 3).14
418 | Carbohydr. Chem., 2012, 38, 416–445

Fig. 2 Surface modifcation in biomaterial design.
Fig. 3 Examples of signalling molecules modulating cell behaviour in bioactive materials for
tissue engineering.
Carbohydr. Chem., 2012, 38, 416–445 | 419

1.2 On the chemical nature of materials
It is sensible to consider today’s optimal meaning of ‘biomaterial’ from two
different perspectives, the first being concerned with the evolution of
materials science and the wide range of materials options that have opened
up during the last decade or so, and the second being the evolution of health
care technologies.
The classical view of a material has been ‘‘a substance of which things are
made.’’ Materials scientists were taught that there were three primary types of
material, metallic, being based on the metallic bond, ceramic, based on ionic
bonds and polymeric, based on covalent bonds. In addition there were hybrids,
which could either be entirely synthetic, usually referred to as composites
which typically would be combinations of ceramics and polymers, or the
natural equivalents of these composites, including bone, wood, and ivory.
Obviously each of these categories contained many subcategories. The metallic
materials included pure metals and alloys, ceramics included glasses, glass–
ceramics and carbons, the polymers included thermosets, thermoplastics,
elastomers and textiles. As biomaterials science emerged, the conventional
view of materials, as being tangible pieces of substances from which useful
objects were made, prevailed. The stems of hip replacements were made of
metals, artificial arteries were made of textiles, dentures and intraocular lenses
were made of acrylic polymers. However, these boundaries between material
classes have now been eroded; those substances derived from clear, chemically
defined primary interatomic and intermolecular bonds are being replaced by
those of greater structural complexity that arise from quite different concepts,
including those of nanotechnology and self assembly processes inspired by
nature. Indeed it is one of the fundamental tenants that is driving nanoscience
and nanotechnology that is at the heart of the revolution in materials science
(or materials chemistry as it is so often called now), and that is the replacement
of top down manufacturing by bottom up synthesis.28
Both biologically derived and synthetic materials have been extensively
explored in regenerative medicine and tissue engineering. In general,
materials from natural sources (e.g., purified protein components such as
collagens from animal tissues) are advantageous because of their inherent
properties of biological recognition, including presentation of receptor-
binding ligands and susceptibility to cell-triggered proteolytic degradation
and remodeling. Despite these advantages, many issues have spurred the
development of synthetic biomaterials as cellular substrates, including
complexities associated with purification, immunogenicity and pathogen
transmission. Although some of these limitations can be overcome by
recombinant protein expression technologies,29 greater control over mate-
rials properties and tissue responses could be achieved if synthetic analogs
were available.30
1.3 Functionalization strategies

Different biomedical devices and applications require different properties
and functions of materials. Therefore methods to modify materials able to
meet the needs of different biomedical systems are extremely variable.
The surfaces of nanostructured materials can be modified and functio-
nalized with different reagents using different methods, including physical,
420 | Carbohydr. Chem., 2012, 38, 416–445

chemical, or biological. Two things are often achieved by surface mod-
ification of materials: 1) enhanced stability and biocompatibility of nano-
structured materials in aqueous media and 2) new material functions and
properties.31
The surface modification of biomaterials with bioactive molecules is a
simple way to make smart materials. To date, different strategies have been
used for the introduction of biomimetic elements into synthetic materials:
a) Physical adsorption (van der Waals, electrostatic, affinity, adsorbed
and cross linked).
b) Physical entrapment attachment (barrier system, hydrogel, dispersed
matrix system).
c) Covalent surface immobilization, taking advantage of different natural
or unnatural functional groups present both on the biomolecules and on the
material surfaces (chemoselective ligation, via amino functionalities, het-
erobifunctional linkers, etc.).
The major methods of immobilizing a bioactive compound to a
polymeric surface are adsorption via electrostatic interactions, ligand–
receptor pairing (as in the couple biotin–avidin), and covalent attachment
(Fig. 4).
Physical methods, such as molecular coating (or adsorption), surface
entrapment, and physical treating with plasma, ozone, or UV have emerged
as leading strategies for surface modifications of materials. Through phy-
sical modifications, functional molecules and entities, varying charges, or
active chemical groups can be introduced onto materials, leading to the
functionalization and bioactivation of materials. Physical modifications
have obvious advantages, including ease of handling and mild interactions
with biomolecules through little or no damage to their bioactivities. These
methods, however, also exhibit certain limitations. For example, physical
interactions are often formed between the substrates and coatings using
these strategies. These physical interactions are usually weak compared to
chemical bonds. Therefore, functional molecules and entities may detach
from the surfaces especially when certain serum compounds compete for
Fig. 4 Different strategies for the introduction of biomimetic elements into synthetic
materials.
Carbohydr. Chem., 2012, 38, 416–445 | 421

active binding sites in physiological conditions; detachment is worsened by
a competitive displacement of theses functional molecules and entities. Non-
covalent adsorption is however sometimes desirable, for example in drug
delivery applications.32
Covalent immobilizations offer several advantages by providing the most
stable bond between the compound and the functionalized polymer surface.
A covalent immobilization can be used to extend the half-life of a biomo-
lecule, prevent its metabolism, or allow continued bioactivity of indwelling
devices.33
Bioactive molecules (growth factors, ECM proteins, etc.) that are free in
solution, as opposed to those immobilized to the matrix, may induce sig-
nificantly different biological responses. Growth factors are routinely added
to cultures in vitro, and have been incorporated and released from polymeric
systems with retention of bioactivity, as shown for neurotrophins, BMPs,
and VEGF.34–36 In vivo, these soluble factors can be released from the
delivery site, and the relevant parameter is the duration over which ther-
apeutic concentrations can be maintained. Alternatively, bioactive mole-
cules can be linked covalently to the scaffold, eventually in a reversible way
or exploiting a degradable linking tether.
These scaffolds have been termed ‘‘cell-responsive’’37 due to release of the
factor upon cellular demand. Once released, these soluble factors can bind
their receptors and initiate the signalling cascade.
Alternatively, immobilized biomolecules can ligate their receptors directly
from the material surface; however, this type of interaction may not exactly
replicate the signalling of soluble factors, as growth factor internalization
can stimulate signalling pathways different from those activated at the
surface. For example, neuronal growth factor NGF induces neurite out-
growth by signalling at the plasma membrane, yet promotes neuron survival
when internalized.38 Surface immobilization has been successfully used to
attach several factors such as EGF,39 BMP-7,40 BMP-2,41 VEGF,42 NGF,43
and NT-344 to a variety of natural and synthetic biomaterials. Signalling by
these immobilized or locally released bioactive ligands may be more potent
than signalling by soluble versions added directly to culture media.45 These
studies also demonstrate that the immobilization strategy must consider
protein structure and active region topology, when designing suitable
delivery systems in order to maximize bioactivity. Ultimately, some factors
may be best delivered in a continuous manner, while others benefit from
direct attachment to the biomaterial surface.46
Different methods have been developed for covalent functionalization of
biomolecules to diverse biomaterials. For covalent functionalization to an
inert solid polymer, first the surface must be modified to provide reactive
groups (-OH, -NH2, COOH, -SH) for the subsequent functionalization step.
When the material does not contain reactive groups, they can be generated
by chemical and physical modification. With this goal, a wide number of
surface modification techniques have been developed, including plasma,47
ionic radiation graft polymerization, photochemical grafting, chemical
modification and chemical derivatization.
In this approach, however, the biomolecules will be linked to the material
surface with a stochastic orientation; this implies that not all the molecules
422 | Carbohydr. Chem., 2012, 38, 416–445

will be exposed efficiently for receptor recognition. Moreover, the chemical
nature of the material can deeply influence the conformation of the mole-
cule,48,49 that can result in an inactive conformation.
For example peptides can react via the N-terminus with different groups
on polymers (Fig. 5). This is usually done by reacting an activated car-
boxylic acid group with the nucleophilic N-terminus of peptides. The car-
boxylic group can be activated with different peptide coupling reagent, e.g.
1-ethyl-3-(3-dimethylaminopropyl)carbodiimide (EDC, also referred to as
water soluble carbodiimide, WSC), dicyclohexylcarbodiimide (DCC) or
carbonyl diimidazole (CDI).
More recent approaches capitalize on chemoselective strategies (Fig. 6),
where selected pairs of functional groups are used to form stable bonds
without the need of an activating agent and without interfering with other
functionalities usually encountered in biomolecules. Chemoselective liga-
tion proceeds usually under mild conditions and results in good yields,50
and allows the linkage of the biomolecule with a controlled spatial
orientation.
A biomolecule may also be attached by these coupling methods via a
spacer group, in order to give better access to the target receptor. One useful
and biocompatible spacer is PEG that has been differently functionalized at
two extremities. Metal or ceramic surface may also be silanized, exploiting
functionalized triethoxysilanes.51
A O O
EDC O N
OH HO N
O O O
O O
B O O O
O O
O
N O ON N ON
NH2 N N
H H H
O O O
C O O
S CF3 S CF3
OH Cl O N
O O H
D O O
O
O HCl O NO2 O O NO2 O N
H
Material
Biomolecule
A) Carboxyl groups, preactivated via EDC and NHS

B) Amino groups, preactivated with DSC
C) Hydroxyl groups, preactivated as tresylate
D) Hydroxyl groups, preactivated as p-nitrophenylcarbonate
Fig. 5 Coupling methods to different groups on materials.
Carbohydr. Chem., 2012, 38, 416–445 | 423

A O
Br
SH S
O
B
H 2N
O O
N
O
H
H
C
HS
O O
S
O O
D O
O
HS
N S
N
E O
O
N3 N
N N
Material/Biomolecule
Biomolecule/Material
A) Thiol and bromoacetyl biomolecule

B) Aldehyde and oxyamine biomolecule
C) Acrylate and thiol biomolecule
D) Maleinimde and thiol biomolecule
E) Azide and propargyl biomolecule
Fig. 6 Examples of chemoselective strategies.
2 Glycoscience for biomaterials design

The glycome, or complement of sugars in the cells, varies among cell type
but also within cell type growing to differences such as stage of differ-
entiation and malignant transformation.52 The traditional view of carbo-
hydrate polymers as energy source (starch and glycogen) and structural
materials (cellulose, collagen, and proteoglycans) has expanded. Today,
carbohydrates are known to have a wide variety of biological functions. For
example, the sulfated polysaccharide, heparin, plays an essential role in
blood coagulation.53 Another related polysaccharide, hyaluronan, which
acts as a lubricant in joints, has been used to protect the corneal endothe-
lium during ophthalmologic surgery.54 In addition to hyaluronan’s lubri-
cating and cushioning properties, polysaccharide and glycoprotein
participate in a number of recognizing processes. Synthetic carbohydrate
based polymers are increasingly being explored as biodegradable, bio-
compatible and biorenewable materials for use as water absorbent, chro-
matographic supports and medical devices. In addition, also membrane
proteins, i.e. lectins, bind specific carbohydrates. This binding is extremely
specific,55 and may present an attractive target for rational design of smart
biomaterials. Exploiting lectin–sugar interactions increases design flexibility
424 | Carbohydr. Chem., 2012, 38, 416–445

by offering the possibility of grafting active sugars on biomaterial surfaces.
One successful exploitation of lectin binding is the creation of galactose-
grafted scaffolds to support hepatocyte (liver cell) growth (see next sec-
tions).56,57 It is notable that the galactose receptor normally functions to
bind soluble sugars, not to mediate adhesion. Thus, it is clear that some
molecules not naturally involved with adhesion may be pressed into that
service through creative engineering. This is not a universal principle,
however, and care must be taken in interpreting results. Most molecules
that bind to a cell surface component (either specifically or non-specifically)
will allow some degree of adhesion.52
Hajime Sato et al. exploited the anionic polysaccharide heparin to extend
the scope of micropatterned carbohydrate displays by applying electrostatic
interaction. They also used micropatterned carbohydrate displays on silicon
substrates by a combination of photolithography and self-assembly.58 These
techniques have interesting potential applications in tissue engineering.
Top-down photolithography and bottom-up molecular self-assembly are
the two main approaches to biomacromolecular micropatterning. Photo-
lithography has been widely used in electronic devices and most practically
applied for precise micropatterning of proteins for cell cultivation.59
Heparin is a glycosaminoglycan polysaccharide with a variable number of
sulfate groups and its binding to basic fibroblast growth factor (bFGF) is
known to modulate cell proliferation and differentiation. The interaction of
heparin with bFGF induces the dimerization of bFGF, which is required for
binding of bFGF to FGF receptor (FGF-R) on cell surfaces,60,61 as illu-
strated below (Fig. 7).
The self-assembled monolayer (SAM) of 3-aminopropyltrimethoxysilane
(APS) as ammonium-terminated template was prepared on silicon and glass
substrates, and micropatterned by photolithography. Self-assembling of
heparin onto the APS-SAM through electrostatic interaction and then
molecular recognition of bFGF to heparin and of fibroblast cells to bFGF
has been achieved.62 Adsorption of heparin was not observed on a silicon
substrate without APS-SAM. Heparin adsorbed on APS-SAM remained on
the substrate after rinsing with PBS buffer. The electrostatic interaction was
essential for anchoring heparin on APS-SAM. The time-course of heparin
adsorption to APS-SAM does not fit with a simple Langmuir model, sug-
gesting that, in addition to electrostatic interaction, some other interaction
Fibroblast
bFGF
Heparin
APS
APS-SAM
Fig. 7 A three-step method for monosaccharide grafting on PET fibers.
Carbohydr. Chem., 2012, 38, 416–445 | 425

such as hydrogen bonding may contribute a little to the adsorption of
heparin to APS-SAM.63,64
The above carbohydrate display using heparin has an attractive potential
for tissue engineering, since it allows a facile and effective strategy for
biomacromolecular patterning with significant bioactive molecules.
In the following sections examples of material functionalization (‘‘bio-
decoration’’) with signalling and relevant glycidic scaffolds will be outlined.
Only recent research concerning the development of smart biomaterials for
Tissue Engineering (TE) applications will be considered here, while gly-
coarrays and glycated nanoparticles for theragnostic application will not be
included. The sections will be organized based on the chemistry used for
covalent attachment of the carbohydrates to the material surfaces.
2.1 Plasma grafting toward ‘‘glyco-decoration’’

Surface modification finds numerous applications in the preparation of
biomaterials, for example, for the control of interfacial interactions between
materials and biological molecules.65–67 Plasma grafting is used in order to
modify the surface features,68 by the introduction of functional groups that
can alter the physico-chemical properties of the surface itself (i.e. hydro-
philicity) and at the same time allow covalent functionalization of the
material with polymers,69 peptides70 or carbohydrates.71 Moreover, this
method is particularly advantageous when the material surface lacks in
functional groups. Extremely high levels of surface functionality can be
obtained by this approach. Examples successfully devised in the past include
amine,72 anhydride,73 epoxide,74 carboxylic acid,75 cyano,76 halide,77
hydroxyl,78 furfuryl,79 and perfluoroalkyl80 functionalized surfaces.
This method brings many advantages in changing the hydrophobicity or
introducing functional groups homogeneously with a defined density over
the inert material surfaces.81 Effectively, any surface that relies on specific
chemistry for its performance can, in principle, be generated by the afore-
mentioned pulsed plasmachemical methodology. Then, friction properties
and adhesion qualities of the treated material can be improved by many
attempts of plasma treatments such as He, Ar or N2.82,83 Plasma treatment
has also the peculiarity of avoiding the use of hazardous solvents without
changing the morphology or the bulk properties of the materials treated.84
Moreover, this method of surface modification can be used in combination
with other well known chemical techniques in order to introduce biomole-
cules in general and carbohydrates in particular in a chemoselective
approach. Two different approaches of the plasma grafting have been used
with the final aim of introducing carbohydrates on the surface of different
polymers. Bech and coworkers85 described an interesting procedure to use
the plasma grafting method in order to functionalize poly(ethylene ter-
ephtalate) (PET) fiber surfaces with monosaccharides. PET is often used in
biomedical material as cardiovascular implants, artificial blood vessels, and
in other applications due to its excellent mechanical properties and fair
biocompatibility.86–88 However, for biomedical applications, the polymer
surfaces are desired to be highly hydrophilic and able to prevent protein
adsorption.89 The modification of polymer fibers to create new bioactive
materials is beneficial for a wide range of applications. For example, the
426 | Carbohydr. Chem., 2012, 38, 416–445

current worldwide crisis in microbial infection and health care of wounds
suggests that more research is necessary to understand practical and effec-
tive ways of creating safe bioactive textiles. The future development of
biomedical and protective textiles with selective properties that benefit the
consumer will be based on applying scientific and clinical advances in
wound healing, antimicrobial, and enzyme-based fabrics.90 In particular,
PET fibers are particularly suited for the design of blood filters due to their
ability to exhibit the highest leukocyte adhesion by the hydrophobic inter-
action between PET and integrins.91 Moreover, PET fibers are often used in
the biomedical field, as implants, vascular grafts for synthetic vessel repla-
cement, or as cardiac heart valves;92 among all synthetic fibers, PET fibers
challenge cotton as the most common textile fiber, due to numerous
advantages, including the cheapness and availability of the raw materials,
the possibility to recycle PET fibers, the melt spinning process used for the
production of PET fibers, which is clean and economical, the relatively high
temperature resistance.
The plasma treatment alone gives surfaces that are relatively unstable
(rapid loss of surface hydrophilicity), so one method is to treat the surface of
PET with appropriate molecules to maintain the surface with a high level of
hydrophilicity and functionality and to achieve a compatible interface with
biological materials. A three steps strategy including two different argon
plasma treatments was successfully used for PET biodecoration with
monosaccharides (Fig. 8). During the first step (pre-activation), the
plasma treatment generated radical species on PET surface that, once
exposed to air, formed peroxide and hydroperoxide reactive groups. In the
second step, the pre-activated fibers were dipped in different mono-
saccharide solutions, specifically, allyl a-D-galactopyranoside (AG) and 2-
methacryloxyethyl glucoside (GEMA). A final plasma treatment allowed
the polymerization of the monosaccharides present on the fiber surface.
This method allows the control of the grafting density by varying the
monosaccharide concentration and the mass of fiber exposed in the plasma
Fig. 8 Plasma grafting of glucose units on PET fibers.
Carbohydr. Chem., 2012, 38, 416–445 | 427

reactor. More interesting results might be obtained with the grafting of
oligosaccharides, which can be selectively recognized by target proteins.
A different strategy has been used by our group in a procedure, com-
bining plasma treatment with click-chemistry technique (Fig. 9). In order
to apply Huisgen cycloaddition to material modification, suitable methods
for azido groups introduction are needed. Several strategies have been
proposed for the preparation of azido-functionalized polymers; among
them we can mention a general strategy consisting in the (co)polymerization
of halide functionalized monomers followed by nucleophilic substitution of
the halide groups by sodium azide (NaN3).93–95 Alternatively, the azido
group can be introduced via epoxide opening by NaN3 on the co-poly-
merized glycidyl-containing monomers (i.e. glycidyl methacrylate).96 A
second approach consists in the (co)polymerization of azido-pre-
functionalized monomers. However, both approaches suffer from a strong
limitation: the necessity of using only particular functionalized monomers
(halide-, epoxide-, azido-functionalized), thus restricting the possibility of
tailoring the chemical and physico-mechanical properties of the material,
which could result, in this way, unsuitable for the desired application.
Plasma treatment was used to introduce amino-groups on polypropylene
(PP) membrane surfaces (preactivation). A diazo-transfer reaction trans-
formed the amino groups into azido groups by triflyl azide solution. Finally
Fig. 9 Diazo-transfer methodology on plasma treated PP surfaces.
428 | Carbohydr. Chem., 2012, 38, 416–445

azido groups were used to introduce by click reaction a propargyl 2-acet-
amido-2-deoxy-a-D-glucopyranoside.
The proposed method is a proof of concepts that opens the way to the
modification of material surfaces lacking reactive functional groups; it
appears an ideal platform for subsequent ‘click’ reactions, that allows the
decoration of the material with any compound, including oligosaccharides,
presenting carbon-carbon triple bonds. The method can be of wide
applicability, allowing the functionalization of materials of different nature
with a wide variety of (bio)active compounds.
2.2 Material functionalization with glycidic structures via amide bond

formation
A different approach capitalize on the amide bond formation between
carboxyl and amino groups suitably located in a complementary position on
the material and on the carbohydrate moieties, respectively.
An example of monosaccharide linkage to materials via amidation was
proposed by our research group on hydroxyapatite.97 Among various
biocompatible materials, synthetic hydroxyapatite (HA) and carbonate
hydroxyapatite (CHA) are widely used in many biomedical applications.98
HA is the natural mineral ingredient of bones, tooth and calcified tissues in
vertebrate. Synthetic HA and CHA are used for human implant coatings
possessing a proven osteoconductivity and its geometry may deeply influ-
ence specific tissue responses, as it happens for vascular in-growth.99
Nonetheless, researchers try to upgrade these inorganic materials by the
addition of biological cues, activating peculiar specific cell-biomaterial
interactions, thereby triggering additional, and possibly more specific bio-
logical responses. Unfortunately HA possess a paucity of reactive func-
tional groups on its surface; therefore covalent linkage of biomolecules on
this material is still a challenging task, despite the observation that the
combination of inorganic materials, such as CHA, and organic signalling
molecules is very promising in tissue engineering.100,101 To date, main
efforts on apatite covalent functionalization focused on the biodecoration of
this inorganic material with whole proteins102 or short peptide epitopes; the
most commonly used peptide for surface modification is RGD,103–105 the
signalling domain derived from fibronectin and laminin.106 Less exploited as
cellular ligands for material functionalization are carbohydrate epitopes
that might allow HA-cell interactions. In this context, the possibility to
decorate hydroxyapatite with carbohydrates can be of relevant interest.
Hence, hydroxyapatite granules (400–600 mm) were treated with amino-
propyltriethoxysilane (APTES) in order to introduce free amino groups on
HA surface (Fig. 10). The free amino functionalities were then condensed
with a carboxyl group introduced onto a C-glycosyl scaffold bearing a
carboxylic group. The method highlighted the possibility to link carbohy-
drates and their analogs to inorganic materials such as HA.
Amide bond formation was also used for the functionalization of chit-
osan as the scaffold with lactobionic acid.107 Among natural biomaterials
chitosan is a binary heteropolysaccharide consisting of (1-4)-linked 2-
acetamido-2-deoxy-b-D-glucopyranose and 2-amino-2-deoxy-b-D-gluco-
pyranose residues,108,109 which has been widely applied in clinic and
Carbohydr. Chem., 2012, 38, 416–445 | 429

Fig. 10 HA functionalization via amide bond formation.
exhibited essential advantage in biodegeneration, biocompatibility and non-

immunoreaction to human body when compared to synthetic polymers. The
aim of this study was to develop a novel natural nanofibrous scaffold with
surface-bound galactose ligands to enhance the bioactivity and mechanical
stability of primary hepatocytes in culture for liver regeneration.110 Primary
hepatocyte culture plays an important role in the clinical treatment of
hepatic failure patients111 to avoid natural human immune response and
provide sufficient replacement of the synthetic and metabolic functions of
the liver, which has been applied in many parts of bioengineering field such
as bioartificial liver-assist device (BLA)112 and tissue engineering for liver
regeneration.113 It is well known that the glycidic part of glycolipids and
glycoproteins strongly interacts with lectins present at cell surfaces:114 In
particular, b-galactose residues are recognized by asialoglycoprotein
receptors on the surfaces of hepatocytes115 and promote hepatocyte
attachment.56,116–120 Thus, many works were concerned with developing
new scaffolds modified with galactose ligands (sse also next section) to
enhance hepatocyte adhesion, maintenance of liver-specific functions and
mechanical stability.121–123
The needed galactose unit was linked to chitosan through the modifica-
tion of lactobionic acid (LA), as represented in Fig. 11; the coupling reac-
tion was performed by an activation step using EDC and NHS.124 The
functionalization of chitosan with galactose improved the adhesion of
hepatocytes on chitosan membrane underlining the bioactive role carried by
the added carbohydrate. In another recent work125 the same bioactive role
of galactose has been exploited in order to differentiate mouse embryonic
stem cells towards hepatocytes. Even in this brilliant work galactose
units have been covalently linked to a p-vinylbenzylamine forming N-p-
vinylbenzyl-O-b-D-galactopyranosyl-(1-4)-D-gluconamide by amidation
reaction.126
Otherwise, the amide reaction can be used to bind carbohydrates to
polysiloxanes (PDMS) in order to obtain ‘‘crew-cut’’ micelles proposed by
many authors.127,128 PDMS could be useful for cell immobilization, medical
tubing, column materials, etc. Most of these research works focused on the
polymers with C–C main chains bearing saccharide residues, with relatively
few chemists paying attention to silicon backbone polymers. Compared
430 | Carbohydr. Chem., 2012, 38, 416–445

Fig. 11 Lactobionic grafting to chitosan via amide bond formation.
with the carbon chain polymers, polysiloxanes have remarkable properties

such as flexibility, low cohesive energy, and biocompatibility.
2.3 Surface grafting by 2,4,6-trichloro-1,3,5-triazine (CY)

Gotoh and coworkers129–131 extensively studied a biomaterial based on a
glycoconjugate lactose-CY-silk fibroin (SF) with the aim to expose galac-
tose units on the material for hepatocyte cultures. Silk fibroin (SF) is a
natural fibrous protein created by the Bombyx mori silkworm, mainly
consisting of the Gly–Ala–Gly–Ala–Gly–Ser sequence having its native
molecular weight of 370 kDa.132 SF fiber has been used as a textile material
for a long time. Recently, it has been found that SF aqueous solution can be
easily converted to gels,133 sponge,134 powder,135 and membranes,136 which
have unique physicochemical properties and biocompatibilities. On the
basis of these observations, SF has been reassessed as biotechnological and
biomedical materials, in particular many current studies on SF explore SF-
based scaffolds for tissue engineering.137 Therefore, it is expected that the b-
galactose non-reducing end of lactose moieties linked to SF may act as an
efficient ligand for hepatocyte attachment to SF scaffolds.
The activated lactose-CY (Fig. 12) was synthesized by the reaction of the
anomeric hydroxyl group with a chlorine atom of CY, then a second
chlorine atom present on the coupling spacer reacts with the phenolic
hydroxyl group of tyrosines residue or with the e-amino group of lysine
residues of the protein sequence. The reaction with the CY allows the
correct exposure of the b-galactose moiety of the disaccharide on the surface
of SF. The covalent linkage of lactose into SF was confirmed by physical
methods and lectins recognition. The positive effect of galactose exposure
on SF was observed by evaluation of the attachment of hepatocytes using
the Lac–CY–SF conjugate-coated dishes, as preliminary tissue enginnering
model.
Carbohydr. Chem., 2012, 38, 416–445 | 431

Fig. 12 Lactose conjugation to SF for the design of hepatocyte-specific biomaterials.
The same chemical methodology was used by Acharya et al.,138 in order

to remodel the properties of SF toward the adhesion of a different kind of
cell, such as fibroblasts with encouraging results.
2.4 Material grafting by click chemistry

A high impact on the glyco-conjugation methodologies has been obtained
by the introduction of the ‘‘click-chemistry’’ approach.139 Although click
reactions could be classified into four different categories, the copper(I)-
catalyzed azide–alkyne cycloaddition (CuAAC) reaction represents the
ground idea of a click reaction.140 Nevertheless, the necessity of a more
‘‘bio-safe’’ environment has acted as a strong driving force for expansion of
432 | Carbohydr. Chem., 2012, 38, 416–445

the repertory of practical click reactions.141 Nowadays, reactions such as the
copper(I)-free azide–alkyne cycloaddition with more reactive alkynes,142 the
radical addition between thiols and alkenes or alkynes,143,144 the Michael
addition of thiols with maleimide, the nucleophilic substitution of the para-
fluoro substituent of pentafluorophenyl groups, and the regular and inverse
electron demand Diels–Alder reactions become to be considered as click
reactions. In addition, click chemistry is becoming explored also by material
scientists both for the processing of bulk materials145,146 and for the func-
tionalization of material surfaces.147,148
In a different approach, Buriak and co-workers149 functionalized a
stainless steel (SS) surface with galactose and N-acetylglucosamine moieties
by a thiol-ene click reaction. Stainless steel has excellent physical properties
such as flexibility and strength, together with appropriate chemical prop-
erties that are fundamental requirements for implanted biomedical devices.
However, SS prostheses and devices are in constant contact with the
aggressive body fluid, and often fail and finally fracture due to corrosion. To
improve the biocompatibility and functionality of unmodified stainless steel
implants, researchers have devoted substantial attention toward controlling
and modifing their surface properties.150,151
The process proposed by Buriak and co-workers is illustrated in Fig. 13.
A thin (sub-15 nm) and uniform silica coatings were first deposited by
atomic layer deposition (ALD) on stainless steel in order to generate a
Fig. 13 Methodology for ‘‘glycodecoration’’ of stainless steel surface.
Carbohydr. Chem., 2012, 38, 416–445 | 433

hydroxy-rich interlayer. These ALD coatings act as a platform on which
alkoxysilane chemistry can take place. After that, the trialkoxysilane deri-
vatives of the monosaccharides GlcNAc and Gal prepared via thiol-ene
coupling were conjugated to the surface (Fig. 13). The carbohydrate moi-
eties bound to the stainless steel surface were then detected via a com-
plementary enzyme-linked lectin assay (ELLA). A key point of the process
is the introduction of a high density of hydroxyl groups on the stainless steel
surface in the first step of the functionalization strategy, in order to allow
efficient formation of controllable organic monolayers through alkoxysilane
coupling reactions.
‘‘Photo-click’ reactions were also developed, affording several additional
advantages, such as spatio-temporal reaction control via focused UV-light
irradiation. An example of this class of reaction is the UV light-promoted
grafting method of a molecule containing a photoreactive aryl azide group
coated on a polymer surface.152,153 In the case of organic azides, upon
photolysis with ultraviolet light, a nitrene radical is generated. The nitrene
radical formed is extremely reactive and can undergo a multitude of reac-
tions, for example, insertion into C-H, N-H, and O-H bonds, addition to
olefins, proton abstraction reactions to give the corresponding amine, and
in the case of aryl azides a number of ring expansion reactions have been
observed. Although the precise mechanism of surface modification is not
fully understood, this method has been demonstrated to be very effective for
a wide range of surfaces. This approach has been used recently to covalently
link carbohydrates to PET fibers by Renaudie et al.154 for example
by incorporating a sialyl-Lewis X saccharide as it is known to interact
specifically with L-selectin.155 Thus a new method for grafting of a new
carbohydrate UV-reactive molecule, an azidophenyl lactamine (AzPhLac),
Fig. 14, was proposed.
The phenylazido group on lactose by reductive amination of lactose with
an aminated phenyl azide molecule, affording the b-D-galactopyranosyl-
(1-4)-1-N-[2-(4-azidopnenylamino)ethylamino]-1-deoxy-D-glucitol (AzPhLac);
the PET fibers were dipped (dip-coating method) in the AzPhLac aqueous
solution, then dried and finally irradiated by UV light for the chemical
fixation of the lactamine moiety via conversion of the phenylazido group
to the highly reactive phenylnitrene. The complete surface modification is
still not fully understood; the proposed mechanism by the authors includes
also the formation of secondary amines. In addition, the highly reactive
phenylnitrene can form covalent bonds not only with the hydrocarbon of
the PET surface but also with linkages of AzPhLac molecules in its
immediate surroundings, leading to the formation of a polymeric multilayer
where the nitrene inserts onto the -OH groups. The method still has to
be improved, since PET surface masking was observed during the UV-
irradiation step.
The group of Bertozzi156 proposed a brilliant example of Copper (I)-
catalyzed azide-alkyne cycloaddition (CuAAC) for the chemoselective
ligation of azide-functionalized pyrene and glycan moieties to the alkyne-
functionalized focal point and chain ends of a dendritic scaffold in order to
‘‘glycodecorate’’ single-walled carbon nanotubes (SWNTs) by ultrasonica-
tion in water. The structural, mechanical, electrical, and optical properties
434 | Carbohydr. Chem., 2012, 38, 416–445

Fig. 14 Grafting by UV-irradiation of azidated lactose to PET fibers.
of single-walled carbon nanotubes (SWNTs) have stimulated considerable

interest in their biological applications,157,158 including biosensing, imaging,
intracellular delivery, and cancer cell targeting. However, use of SWNTs in
living systems requires strategies to diminish their cytotoxicity.159 Thus,
surface modifications that lower the toxicity of SWNTs while simulta-
neously enabling specific biological recognition are highly desirable.160,161
A promising approach is to coat SWNTs with synthetic glycopolymers able
to mimic the glycoproteins found on cell surfaces.
The proposed approach involves first the synthesis of dendritic structure
ensuring the mimicry of glycoproteins and a suitable moiety for the anchoring
to the nanotube surface. Thus, a pyrene tail was conjugated to the focal point
of the dendrimer by CuAAC (Fig. 15). The resulting second generation
dendrimer was further coupled with pent-4-ynoic anhydride to introduce
additional alkyne groups. This third generation dendrimer was then reacted
with a 2-azidoethyl mono- or disaccharide using CuAAC once again.
SWNTs coating with glycodendrimers resulted in a bioactive surface that
also mitigate their cytotoxicity. The synthetic method proposed by Bertozzi
can be readily adapted to ligands for other receptor interactions, in view of
novel applications such as biosensors for carbohydrate-binding proteins
and delivery agents that target specific cell-surface receptors.
Carbohydr. Chem., 2012, 38, 416–445 | 435

Fig. 15 Generation of glycodendrimers for carbon nanotubes functionalization.
436 | Carbohydr. Chem., 2012, 38, 416–445

OH O
O
O N3
O N3 O O
HO O
HA granules HA-N3
OH
HO O Huisgen cycloaddition
HO
HO N
O N O
N
O
O O
HA-Glc
Fig. 16 HA functionalization by Huisgen cycloaddition.
The Huisgen-type cycloaddition was also proposed for the glycodecora-

tion of hydroxyapatite (HA); in this example, a minimal spacer of difunc-
tional PEG was used (Fig. 16), possessing an azido group at one term for the
Huisgen cycloaddition, and a carboxyl group on the other side for
the covalent bonding to the hydroxyl groups present on the HA surface.162
The Huisgen cycloaddition was performed onto HA-N3 with propargyl
a-D-glucopyranoside. The ‘‘glycosylated’’ hydroxypatite was characterized
by its ability to interact with glucose recognizing lectins.
Xue-Long Sun and co-workers163 demonstrated the applicability of
sequential Diels-Alder and azide-alkyne [3 þ 2] cycloaddition reaction for
the immobilization of carbohydrates onto a glass-slide surface. A poly-
ethyleneglycol (PEG) linker carrying both an alkyne and a cyclopentadiene
terminal groups was synthesized for the sequential click-reactions. The
commercially available N-(e-maleimidocaproyl) (EMC)-functionalized glass
slide was used as the starting material bearing the dienophile on the material
surface (Fig. 17). Alternatively, other groups have developed useful new
bioorthogonal cycloaddition with different partner pairs, including azide-
oxanorbornadiene,164 nitrone cyclooctyne165 and an extraordinarily rapid
tetrazine–strained alkene reaction,166 but none of these has been applied to
carbohydrate functionalization of material surfaces.
Crucial to the selection of Diels-Alder cycloaddition as a preferential
strategy for surface immobilization of biomolecules are the following key
issues:
a) water is the best solvent for biomolecules;
b) water has an extraordinary rate-accelerating effect on the reaction
process;
c) the reaction occurs efficiently at room temperature;
d) the diene and dienophiles of choice (cyclopentadiene and EMC) are
independently stable and easy to introduce.
The first reaction takes place between the difunctional PEG and the
maleimido group of the glass surface, affording the corresponding
Carbohydr. Chem., 2012, 38, 416–445 | 437

Fig. 17 Double click reaction for the functionalization of glass slides.
438 | Carbohydr. Chem., 2012, 38, 416–445

Fig. 18 SL conjugation method to chitosan-based materials.
unsaturated six-membered ring (Fig. 17, step 1). This reaction occurs at
room temperature without any influence by solvent and pH. In order to bio-
functionalize the PEG surface with carbohydrates, a second step involving a
Huisgen type cycloaddition with a suitable azidolactoside (Fig. 17, step 2)
was performed. The process efficacy of the coupled click reactions was
determined with FITC-galectin and with the use of azidated biotin as the
alkyne partner in the second step.
2.5 Miscellaneous grafting methods

A very intersting and up-to-date example is given by the functionalization
with carbohydrates of chitosan-based material proposed as anti-viral
agents. In fact, the coupling of glycoligands to material surfaces may gen-
erate appealing multivalent systems that could be used as diagnostic or
therapy tools (theragnostics).
Influence virus start the infection of host cells by its surface hemagglu-
tinin (HA) protein that binds to its glycoligands, such as sialyllactose (SL).
Carbohydr. Chem., 2012, 38, 416–445 | 439

The group of Cheng and co-workers167 presented the construction of two
SL-incorporated chitosan-based materials, either as a water-soluble poly-
mer or as a functional fiber, and demonstrated their abilities for viral
adhesion inhibition and decontamination.
The syntheses were accomplished by grafting a lactoside bearing an
aldehyde-functionalized aglycone to the amino groups of chitosan or chit-
osan fiber followed by the enzymatic sialylation with sialyltransferase
(Fig. 18). The obtained water-soluble SL chitosan conjugate was shown to
bind hemagglutinin with high affinity and inhibited effectively the viral
attachment to host erythrocytes. Moreover, the SL functionalized chitosan
fiber was able to efficiently remove the virus from an aqueous medium. The
results collectively demonstrate that these potential new materials may
function as the virus adsorbents for prevention and control of influenza. In
addition, this example represents an appealing approach for presenting a
protein ligand on a chitosan backbone, which is a versatile molecular
platform for biofunctionalization and, thereby, can be used for not only
antiviral design, but also extensive medical development such as diagnosis,
drug delivery and biomaterial for tissue engineering applications.
Conclusions
Few examples of materials biofunctionalized with carbohydrates have
appeared recently in the literature with the aim of ameliorating materials
features, in terms of bioactivity, biocompatibility, and biostability. Despite
the increasing development of glycomics and the methods at our disposal
for oligossacharide synthesis, application of glycoscience for the design of
biomaterials for tissue engineering is still scarce. We think that in the next
few years a great contribution by glycochemists will be given to this research
field, that may experience a great explosion as observed for glycosylated
nanoparticles and glycoarrays.
References
1 N. Huebsch and D. J. Mooney, Nature, 2009, 462, 426.
2 M. J. Lysaght, A. Jaklenec and E. Deweerd, Tissue Eng Part A, 2008, 14, 305.
3 US Department of Health and Human Services. 2020: A New Vision - A Future
for Regenerative Medicine, 2006, http://www.hhs.gov/reference/newfuture.shtml
4 Bouchie, Nature Biotechnol, 2002, 20, 1178.
5 E. S. Place, N. D. Evans and M. M. Stevens, Nature Mat, 2009, 8, 457.
6 B. D. Ratner and S. J. Bryant, Annu Rev Biomed Eng, 2004, 6, 41.
7 J. M. Anderson, A. Rodriguez and D. T. Chang, Semin Immunol, 2008, 20, 86.
8 G. Balasundaram and T. J. Webster, J Mater Chem, 2006, 16, 3737.
9 S. C. G. Leeuwenburgh, J. A. Jansen, J. Malda, A. D. Wouter, J. Rouwkema,
C. A. Van Blitterswijk, C. J. Kirkpatrick and D. F. Williams, Biomaterials,
2008, 29, 3047.
10 R. Langer and D. Tirrell, Nature, 2004, 428, 487.
11 A. Hubbel, J. Patterson and M. Martino, Materials Today, 2010, 13, 14.
12 U. Mayer and R. Timpl, Nidogen: A versatile binding protein of basement
membranes. In Extracellular Matrix Assembly and Structure, 1994, (P. D.
Yurchenco, D. E. Birk and R. P. Mecham eds.) Academic Press, San Diego,
pp. 318–416.
440 | Carbohydr. Chem., 2012, 38, 416–445

13 J. C. M. van Hest and D. A. Tirrell, Chem Commun, 2001, 19, 1897.
14 W. P. Daley, S. B. Peters and M. Larsen, J Cell Sci, 2008, 121, 255.
15 N. Zagris, 2001. Micron: Extracellular matrix in development of the early
embryo (Oxford, England, 1993), 32, pp. 427–438.
16 D. Gullberg and P. Ekblom, Int J Dev Biol, 1995, 76, 845.
17 J. E. Scott, J Ant, 1995, 187, 259.
18 M. Aumailley and B. Gayraud, J Mol Med, 1998, 76, 253.
19 E. I. Wallner, Q. Yang, D. R. Peterson, J. Wada and Y. S. Kanwar, Am J
Physiol, 1998, 275, 467.
20 D. J. Behonick and Z. Werb, Mech Dev, 2003, 120, 1327.
21 M. M. Stevens and J. H. George, Science, 2005, 310, 1135.
22 L. Bacáková, E. Filova, F. Rypacek, V. Svorcik and V. Stary, Physiol Res,
2004, 53, 35.
23 J. Gao, Y. M. Kim, H. Coe, B. Zern, B. Sheppard and Y. Wang, PNAS, 2006,
103, 16681.
24 V. H. Fan, A. Au, K. Tanama, R. Littrell, L. B. Richardson, J. W. Wright,
A. Wells and L. G. Griffith, Stem Cells, 2007, 25, 1241.
25 A. Sahni, L. A. Sporn and C. W. Francis, J Biol Chem, 1999, 274, 14936.
26 N. Gomez, Y. Lu, S. Chen and C. E. Schmidt, Biomaterials, 2007, 28,
271.
27 K. Alberti, R. E. Davey, K. Onish, S. George, K. Salchert, F. P. Seib,
M. Bornhauser, T. Pompe, A. Nagy, C. Werner and P. W. Zandstra,
Nat Methods, 2008, 5, 645.
28 D. F. Williams, Biomaterials, 2009, 30, 5897.
29 C. Yang, P. J. Hillas, J. A. Baez, M. Nokelainen, J. Balan, J. Tang, R. Spiro
and J. W. Porarek, BioDrugs, 2004, 18, 103.
30 M. P. Lutolf and J. A. Hubbell, Nature Biotechnol., 2005, 23, 47.
31 T. Xu, N. Zhang, H. L. Nichols, D. Shi and X. Wen, Mater Sci Eng C, 2007,
27, 578.
32 K. Langer, C. Coester, C. Weber, H. von Briesen and J. Kreuter, Eur J Pharm
Biopharm, 2000, 49, 303.
33 J. M. Goddard and J. H. Hotchkiss, Prog Pol Sci, 2007, 32, 698.
34 K. J. Whittlesey and L. D. Shea, Exp Neurol, 2004, 109, 1.
35 C. A. Simmons, E. Alsberg, S. Hsiong, W. J. Kim and D. J. Mooney, Bone,
2004, 35, 562.
36 H. C. Kroese-Deutman, P. Q. Ruhe, P. H. Spauwen and J. A. Jansen,
37 M. P. Lutolf, G. P. Raeber, A. H. Zisch, N. Tirelli and J. A. Hubbell, Adv
Mater, 2003, 20, 5671.
38 Y. Zhang, D. B. Moheban, B. R. Conway, A. Bhattacharyya and R. A. Segal,
J Neurosci, 2000, 20, 5671.
39 P. R. Kuhl and L. G. Griffith-Cima, Nat Med, 1996, 2, 1022.
40 K. Kirkwood, B. Rheude, Y. J. Kim, K. White and K. C. Dee, J Oral
Implantol, 2003, 29, 57.
41 V. Karageorgiou, L. Meinel, S. Hofmann, A. Malhotra, V. Volloch and D.
Kaplan, J Biomed Mater Res A, 71, 528.
42 A. H. Zisch, M. P. Lutolf, M. Ehrbar, G. P. Raeber, S. C. Rizzi, N. Davies, H.
Schmökel, D. Bezuidenhout, V. Djonov, P. Zilla and J. A. Hubbel, FASEB J,
2003, 17, 2260.
43 A. C. Lee, V. M. Yu, J. B. Lowe, M. J. Brenner, D. A. Hunter, S. E.
Mackinnon and S. E. Sakiyama-Elbert, Exp Neurol, 2003, 184, 295.
44 S. J. Taylor, J. W. McDonald and S. E. Sakiyama-Elbert, J Control Relase,
2004, 98, 281.
Carbohydr. Chem., 2012, 38, 416–445 | 441

45 D. Seliktar, A. H. Zisch, M. P. Lutolf, J. L. Wrana and J. A. Hubbell,
J Biomed Mater Res A, 2004, 68, 704.
46 I. D. Dinbergs, L. Brown and E. R. Edelman, J Biol Chem, 1996, 271, 29822.
47 L. Russo, S. Zanini, C. Riccardi, F. Nicotra and L. Cipolla, Mat Today, 2011,
14, 164.
48 G. Goobes, P. S. Stayton and G. P. Drobny, Progr. Nuclear Magn. Res.
Spectr., 2007, 50, 71.
49 G. Goobes, R. Goobes, W. J. Shaw, J. M. Gibson, J. R. Long, V.
Raghunathan, O. Schueler-Furman, J. M. Popham, D. Baker, C. T. Campbell,
P. S. Stayton and G. P. Drobny, Magn. Reson. Chem., 2007, 45, S32.
50 U. Hersel, C. Dahmen and H. Kessler, Biomaterials, 2003, 24, 4385.
51 P. Dubruel, E. Vanderleyden, M. Bergadà, I. De Paepe, H. Chen, S. Kuypers,
J. Luyten, J. Schrooten, L. Van Hoorebeke and E. Schacht, Surface Science,
2006, 600, 2562.
52 M. D. Mager, V. LaPointe and M. M. Stevens, Nature Chem, 2011, 3, 582.
53 R. J. Linhardt, T. Toida, Carbohydrates in Drug Design, 1997, (Witczak, Z. J.,
Nieforth, K. A., eds.), Marcel Dekker: New York, pp. 277–341.
54 E. A. Balazs, T. C. Laurent, The Chemistry, Biology and Medical applications
of hyaluronan and its derivatives, 1998, (Laurent, T. C., ed.), London, Portland
Press, pp. 325–336.
55 W. I. Weis and K. Drickamer, Annu Rev Biochem, 1996, 65, 441.
56 J. A. Oka and P. H. Weigel, J Cell Biol, 1986, 103, 1055.
57 L. G. Griffith and S. Lopina, Biomaterials, 1998, 19, 979.
58 Y. Miura, H. Sato, T. Ikeda, H. Sugimura, O. Takai and K. Kobayashi,
Biomacromolecule, 2004, 5, 1708.
59 C. Love, J. L. Ronan, G. M. Grotenbreg, A. G. van der Veen and H. L.
Ploegh, Nat Biotechnol, 2006, 24, 703.
60 Y. Umezewa and H. Aoki, Anal Chem, 2004, 76, 320.
61 U. Lindahl, K. Lidholt, D. Spillmann and L. Kjellèn, Thrombosis Res, 1994,
75, 1.
62 H. Sato, Y. Miura, N. Saito, K. Kobayashi and O. Takai, Surface Science,
2007, 601, 3871.
63 A. B. Foster, Ann Rev Biochem, 1961, 30, 45.
64 H. Verli and J. A. Guimaraes, Carbohydr Res, 2004, 339, 281.
65 K. S. Siow, L. Britcher, S. Kumar and K. J. Griesser, Plasma Process Polym,
2006, 3, 392.
66 F. Poncin-Epaillard, G. J. Legeay, Biomater Sci Polym Ed, 2003, 14, 1005.
67 J. Wang, C. J. Pan, N. Huang, H. Sun, P. Yang, Y. X. Leng, J. Y. Chen,
G. J. Wan and P. K. Chu, Surface and Coatings Technology, 2005, 196, 307.
68 K. Kato, E. Uchida, Y. Uyama and Y. Ikada, Prog Polym Sci, 2003, 28,
209.
69 C. Oehr, M. Müller, B. Elkin, D. Hegemann and U. Vohrer, Surf. Coat.
Technol., 1999, 3, 25.
70 J. Wang, C. J. Pan, N. Huang, H. Sun, P. Yang, Y. X. Leng, J. Y. Chen, G. J.
Wan and P. K. Chu, Surface and Coatings Technology, 2005, 196, 307.
71 S. Hu, X. Ren, M. Bachman, C. E. Sims, G. P. Li and N. Allbritton, Anal
Chem, 2002, 74, 4117.
72 G. Oye, V. Roucoules, L. J. Oates, A. M. Cameron, N. R. Cameron, P. G.
Steel, B. G. Davis, D. M. Coe, R. A. Cox and J. P. S. J. Badyal, Phys Chem B,
2003, 107, 3496.
73 M. E. Ryan, A. M. Hynes and J. P. S. Badyal, Chem Mater, 1996, 8, 37.
74 C. Tarducci, E. Kinmond, S. Brewer, C. Willis and J. P. S. Badyal, Chem
Mater, 2000, 12, 1884.
442 | Carbohydr. Chem., 2012, 38, 416–445

75 S. J. Hutton, J. M. Crowther and J. P. S. Badyal, Chem Mater, 2000, 12,
2282.
76 C. Tarducci, W. C. E. Schofield, S. Brewer, C. Willis and J. P. S. Badyal, Chem
Mater, 2001, 13, 1800.
77 D. O. H. Teare, D. C. Barwick, W. C. E. Schofield, R. P. Garrod, L. J. Ward
and J. P. S. Badyal, Langmuir, 2005, 21, 11425.
78 C. Tarducci, W. C. E. Schofield, S. A. Brewer, C. Willis and J. P. S. Badyal,
Chem Mater, 2002, 14, 2541.
79 C. Tarducci, S. A. Brewer, C. Willis and J. P. S. Badyal, Chem Commun, 2005,
3, 406.
80 S. R. Coulson, I. S. Woodward, S. A. Brewer, C. Willis and J. P. S. Badyal,
Langmuir, 2000, 16, 6287.
81 R. Q. Kou, Z. K. Xu, H. T. Deng, Z. M. Liu, P. Seta and Y. Y. Xu, Langmuir,
2003, 19, 6869.
82 D. Hegemann, H. Brunner and C. Oehr, Nuclear Instruments and Methods in
Physics Research Section B: Beam Interactions with Materials and Atoms, 2003,
208, 281.
83 J. López Garcı́a, A. Asadinezhad, J. Pachernı́k, M. Lehocký, I. Junkar, P.
Humpolı́ček, P. Sáha and P. Valášek, Molecules, 2010, 15, 2845.
84 K. S. Siow, L. Britcher, S. Kumar and H. J. Griesser, Plasma Process Polym,
2006, 3, 392.
85 L. Bech, B. Lepoittevin, A. Achhab, E. Lepleux, L. Teulè-Gay, C. Boisse-
Laporte and P. Roger, Langmuir, 2007, 23, 10348.
86 G. Legeay, F. Poncin-Epaillard and C. R. Arciola, Int J Artif Organs, 2006,
29, 453.
87 K. Sugiyama, K. Kato, M. Kido, K. Shiraishi, K. Ohga, K. Okada and O.
Matsuo, Macromol Chem Phys, 1998, 199, 1201.
88 M. W. Huh, I. K. Kang, D. H. Lee, W. S. Kim, D. H. Lee, L. S. Park, K. E.
Min and K. H. J. Seo, J Appl Polym Sci, 2001, 81, 2769.
89 H. T. Deng, Z. K. Xu, Z. W. Dai, J. Wu and P. Seta, Enzyme Microbiol
Technol, 2005, 36, 996.
90 G. Sun, Bioactive Fibres and Polymers, 2001, (J. V. Edwards and T. L. Vigo,
eds.), 792, American Chemical Society, Washington, D.C., pp. 243–252.
91 G. Reese, In Modern Polyesters: Chemistry and Technology of Polyesters
and Copolyesters, 2003, (J. Scheirs, T. E. Long, eds.), Wiley, Chichester, pp.
401–433.
92 L. Barbe, B. Boval, M.-P. Wautier and J.-L. Wautier, Transfusion, 2000, 40,
1250.
93 R. T. Chen, B. W. Muir, G. K. Such, A. Postma, R. A. Evans, S. M. Pereira,
K. M. McLean and F. Caruso, Langmuir, 2010, 26, 3388.
94 J. F. Lutz, H. G. Börner and K. Weichenhan, Macromol Rapid Commun, 2005,
26, 514.
95 R. S. Lee and Y. T. Huang, J Polym Sci Part A: Polym Chem, 2008, 46, 4320.
96 N. V. Tsarevsky, S. A. Bencherif and K. Matyjaszewski, Macromolecules,
2007, 40, 4439.
97 M. Sandri, A. Natalello, D. Bini, L. Gabrielli, L. Cipolla and F. Nicotra,
Synlett, 2011, 13, 1845.
98 S. V. Dorozhkin, Biomaterials, 2010, 31, 1465.
99 V. Karageorgiou and D. Kaplan, Biomaterials, 2005, 26, 5474.
100 L. L. Hench and J. M. Polak, Science, 2002, 295, 1014.
101 H. S. Mansur and H. S. Posta, Chem Eng J, 2007, 137, 72.
102 K. Zurlinden, M. Laub and H. P. Jennissen, Materialwiss Werkst, 2005,
36, 820.
Carbohydr. Chem., 2012, 38, 416–445 | 443

103 G. Balasundaram, M. Sato and T. J. Webster, Biomaterials, 2006, 27, 2798.
104 M. Durrieu, S. Pallu, F. Guillemot, R. Bareille, J. Amédée and C. H. Baquey,
J Mater Sci-Mater M, 2004, 15, 779.
105 M. Nelson, M. Balasundaram and T. J. Webster, Int J Nanomed, 2006, 1, 339.
106 U. Hersel, C. Dahmen and H. Kessler, Biomaterials, 2003, 24, 4385.
107 Z. Q. Feng, X. Chu, N. P. Huang, T. Wang, Y. Wang, X. Shi, Y. Ding and
Z. Z. Gu, Biomaterials, 2009, 30, 2753.
108 B. T. Houseman and M. Mrksich, Chem Biol, 2002, 9, 443.
109 B. T. Houseman, E. S. Gawalt and M. Mrksich, Langmuir, 2003, 19, 1522.
110 H. Lis and N. Sharon, Chem Rev, 1998, 98, 637.
111 J. K. Park and D. H. Lee, J Biosci Bioeng, 2005, 99, 311.
112 A. J. Strain and J. M. Neuberger, Science, 2002, 295, 1005.
113 I. K. Kang, J. S. Moon, H. M. Jeon, W. Meng, Y. I. Kim, Y. J. Hwang and S.
Kim, J Mater Sci Mater Med, 2005, 16, 533.
114 A. Varki, Glycobiology, 1993, 3, 97.
115 G. Ashwell and J. Harford, Ann Rev Biochem, 1982, 51, 531.
116 P. H. Weigel, R. L. Schnaar, M. S. Kuhlenschmidt, E. Schmell, R. T. Lee,
Y. C. Lee and S. Roseman, J Biol Chem, 1979, 254, 10830.
117 A. Kobayashi, T. Akaike, K. Kobayashi and H. Sumitomo, Chem Rapid
Commun, 1986, 7, 645.
118 A. Kobayashi, M. Goto, K. Kobayashi and T. Akaike, J Biomater Sci Polym,
1994, 6, 325.
119 J. Yang, M. Goto, H. Ise, C.-S. Cho and T. Akaike, Biomaterials, 2002,
23, 471.
120 J. J. Yoon, Y. S. Nam, J. H. Kim and T. G. Park, Biotechnol Bioeng, 2002,
78, 1.
121 K. A. Peterlinz, R. M. Georgiadis, T. M. Herne and M. J. J. Tarlov, Am Chem
Soc, 1997, 119, 3401.
122 M. Schena, D. Shalon, R. W. Davis and P. O. Brown, Science, 1995, 270, 467.
123 R. Georgiadis, K. P. Peterlinz and A. W. Peterson, J Am Chem Soc, 2000, 122,
3166.
124 I. K. Park, J. Yang, H. J. Jeong, H. S. Bom, I. Harada, T. Akaike, S. I. Kim
and C. S. Cho, Biomaterials, 2003, 24, 2331.
125 Q. Meng, A. Haque, B. Hexiq and T. Akaike, Biomaterials, 2012, 33, 1414.
126 A. Kobayashi, K. Kobayashi and T. Akaike, J Biomater Sci Polym Edn, 1992,
3, 499.
127 A. Halperin, M. Tirrell and T. P. Lodge, Adv Polym Sci, 1992, 100, 31.
128 L. Zhang and A. Elisenberg, Science, 1995, 268, 1728.
129 Y. Gotoh, M. Tsukada, S.-I. Aiba and N. Minoura, Int J Biol Macromol, 1996,
18, 19.
130 Y. Gotoh, N. Minoura and T. Miyashita, Colloid Polym Sci, 2002, 280, 562.
131 Y. Gotoh, S. Niimi, T. Hayakawa and T. Miyashita, Biomaterials, 2004, 25,
1131.
132 C. M. Mello, K. Senecal, B. Yeung, P. Vouros, D. Kaplan, Silk polymers:
materials science and biotechnology, 1994, (D. Kaplan, W. W. Adams,
B. Farmer and C. Viney eds), American Chemical Society, Washington DC,
pp. 67–79.
133 T. Hanawa, A. Watanabe, T. Tsuchiya, R. Ikoma, M. Hidaka and M.
Sugihara, M, Chem Pharm Bull, 1995, 43, 284.
134 M. Li, S. Lu, Z. Wu, H. Yan, J. Mo and L. Wang, J Appl Polym Sci, 2001, 79,
2185.
135 D. Akiyama, Y. Kitahara, L. Xuan, M. Murakami, M. Arai and K.
Hirabayashi, J Soc Fiber Sci Tec, 1991, 47, 339.
444 | Carbohydr. Chem., 2012, 38, 416–445

136 N. Minoura, M. Tsukada and M. Nagura, Biomaterials, 1990, 11, 430.
137 S. Sofia, M. B. McCarthy, G. Gronowicz and D. L. Kaplan, J Biomed Mater
Res, 2001, 54, 139.
138 C. Acharaya, B. Hinz and S. C. Kundu, Biomaterials, 2008, 29, 4665.
139 C. Remzi Becer, R. Hoogenboom and U. S. Schubert, Angew Chem Int Ed,
2009, 48, 4900.
140 H. C. Kolb, M. G. Finn and K. B. Sharpless, Angew Chem Int Ed, 2001, 40,
2004.
141 E. M. Sletten and C. R. Bertozzi, Angew Chem Int Ed, 2009, 48, 6974.
142 M. F. Debets, C. W. J. van der Doelen, F. P. J. T. Rutjes and F. L. van Delft,
ChemBioChem, 2010, 11, 1168.
143 A. Dondoni, Angew Chem Int Ed, 2008, 47, 8995.
144 R. Hoogenboom, Angew Chem Int Ed, 2010, 49, 3415.
145 M. Heggli, N. Tirelli, A. Zisch and A. Hubbell, Bioconjugate Chem, 2003,
14, 967.
146 D. Fournier, R. Hoogenboom and U. S. Schubert, Chem Soc Rev, 2007, 36,
1369.
147 F. L. van Delft, J. C. M. van Hest, R. J. M. Nolte, A. E. Rowan and F. P. J. T.
Rutjes, QSAR, Comb Sci, 2007, 26, 125.
148 F. Santoyo-Gonzalez and F. Hernandez-Mateo, Chem Soc Rev, 2009, 38,
3449.
149 A. M. Slaney, V. A. Wright, P. J. Meloncelli, K. D. Harris, L. J. West, T. L.
Lowary and J. M. Buriak, ACS App Mat Inter, 2011, 3, 1601.
150 G. Sydow-Plum and M. Tabrizian, Mater Sci Technol, 2008, 24, 1127.
151 P. Thevenot, W. Hu and L. Tang, Curr Top Med Chem, 2008, 8, 270.
152 M. A. Harmer, Langmuir, 1991, 7, 2010.
153 T. Sugawara and T. Matsuda, Macromolecules, 1994, 27, 7809.
154 L. Renaudie, C. L. Narvor, E. Lepleux and P. Roger, Biomacromolecules,
2007, 8, 679.
155 J. E. Puksas and Y. Chen, Biomacromolecules, 2004, 5, 1141.
156 P. Wu, X. Chen, N. Hu, U. C. Tam, O. Blixt, A. Zetti and C. R. Bertozzi,
Angew Chem Int Ed, 2008, 47, 5022.
157 D. Tasis, N. Tagmatarchis, A. Bianco and M. Prato, Chem Rev, 2006, 106,
1105.
158 R. S. Kane and A. D. Stroock, Biotechnol Prog, 2007, 23, 316.
159 V. L. Colvin, Nat Biotechnol, 2003, 21, 1166.
160 X. Chen, G. S. Lee, A. Zettl and C. R. Bertozzi, Angew Chem Int Ed, 2004, 43,
6111.
161 X. Chen, U. C. Tam, J. L. Czlapinski, G. S. Lee, D. Rabuka, A. Zettl and
C. R. Bertozzi, J Am Chem Soc, 2006, 128, 6292.
162 L. Russo, E. Landi, A. Tampieri, A. Natalello, S. M. Doglia, L. Gabrielli, L.
Cipolla and F. Nicotra, Carbohydrate Research, 2011, 346, 1564.
163 X. L. Sun, Bioconjugate Chem, 2006, 17, 52.
164 S. S. van Berkel, A. T. Dirks, M. F. Debets, F. L. van Delft, J. J. Cornelissen,
R. J. Nolte and F. P. Rutjes, Chembiochem, 2007, 8, 1504.
165 C. S. McKay, J. Moran and J. P. Pezacki, Chem Commun, 2010, 46, 931.
166 M. L. Blackman, M. Royzen and J. M. Fox, J Am Chem Soc, 2008, 130, 13518.
167 X. Li, P. Wu, G. F. Gao and S. Cheng, Biomacromolecules, 2011, 12, 3962.
Carbohydr. Chem., 2012, 38, 416–445 | 445

Carbohydrate Chemistry

Uploaded by

Copyright:

Available Formats

Carbohydrate Chemistry

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Carbohydrate Chemistry

Uploaded by

Copyright:

Available Formats

Specialist Periodical Reports

Edited by A Pilar Rauter

& The Royal Society of Chemistry 2012

All rights reserved

Published by The Royal Society of Chemistry,

Registered Charity Number 207890

For further information see our web site at www.rsc.org

Carbohydrate research plays a remarkable role in chemistry and biology

Carbohydr. Chem., 2012, 38, vii–viii | vii

viii | Carbohydr. Chem., 2012, 38, vii–viii

Applications of glycobiology: biological and immunological eﬀects of a 1

Lipopolysaccharide structure and biological activity from the cystic 13

Carbohydr. Chem., 2012, 38, ix–xiv | ix

Synthesis of bacterial carbohydrate surface structures containing Kdo 40

Synthetic glycopeptides in vaccine development and antibody epitope 61

Posttranslational sialylation and its impact on leukocyte recruitment 75

x | Carbohydr. Chem., 2012, 38, ix–xiv

Congenital Disorders of Glycosylation (CDG): from glycoproteins 124

Bladder cancer–glycosylation insights 156

Levansucrases of Pseudomonas bacteria: novel approaches for protein 176

Carbohydr. Chem., 2012, 38, ix–xiv | xi

Recent advances on the application of NMR methods to study the 192

Glycosidase inhibitors: versatile tools in glycobiology 215

xii | Carbohydr. Chem., 2012, 38, ix–xiv

Multivalent glycoconjugates in medicinal chemistry 303

Glycotransporters for gene delivery 338

Furanose-based templates in the chemoselective generation of 376

Carbohydr. Chem., 2012, 38, ix–xiv | xiii

Synthesis of carbohydrate-based artiﬁcial siderophores and their 398

Smart biomaterials: the contribution of glycoscience 416

xiv | Carbohydr. Chem., 2012, 38, ix–xiv

Carbohydrate chemistry, oligosaccharide sequencing, synthesis technolo-

Carbohydr. Chem., 2012, 38, 1–12 | 1

2 Historical breakthroughs and examples

2 | Carbohydr. Chem., 2012, 38, 1–12

Carbohydr. Chem., 2012, 38, 1–12 | 3

4 | Carbohydr. Chem., 2012, 38, 1–12

3 Polysaccharides and derivatives

4 An historical ﬁnding in virology?

Carbohydr. Chem., 2012, 38, 1–12 | 5

polyanions, such as polymethacrylates and polyacrylates, as antiviral agents

5 COAM does not induce interferon

6 | Carbohydr. Chem., 2012, 38, 1–12

dose-dependently the cell cultures against mengovirus, but only poly IC

Carbohydr. Chem., 2012, 38, 1–12 | 7

Coxsackie B4 Mice (profyl. 24 h) 2 mg/mouse ip. Protection 53

7 Therapeutic implications for acute and chronic inﬂammation and cancer

8 | Carbohydr. Chem., 2012, 38, 1–12

8 Conclusions and future perspectives

Carbohydr. Chem., 2012, 38, 1–12 | 9

10 | Carbohydr. Chem., 2012, 38, 1–12

Carbohydr. Chem., 2012, 38, 1–12 | 11

12 | Carbohydr. Chem., 2012, 38, 1–12

Carbohydr. Chem., 2012, 38, 13–39 | 13

1.2 Cystic Fibrosis as an inﬂammatory disorder

14 | Carbohydr. Chem., 2012, 38, 13–39

1.3 Cystic Fibrosis as a complex microbial community disease

Carbohydr. Chem., 2012, 38, 13–39 | 15

2 Basic Lipopolysaccharide structure

Fig. 1 Scheme of a general chemical structure of bacterial lipopolysaccharides.